Document 13313560

advertisement
Leucine Rich Repeat Kinase 2 in the pathogenesis of Parkinson’s Disease
by
Tatiana D. Papkovskaia
A thesis submitted for the degree of Doctor of Philosophy
Department of Clinical Neuroscience, Institute of Neurology
University College London
2013
1
Declaration
I, Tatiana D. Papkovskaia confirm that the work presented in this thesis is my own.
The help and contribution of others to this thesis is specified in the
acknowledgements section. Where information has been derived from other
sources, I confirm that this has been indicated in the thesis.
2
To my parents
3
Abstract
The Leucine Rich Repeat kinase 2 (LRRK2) G2019S mutation is the most
common genetic cause of Parkinons's disease (PD) which is clinically and
pathologically indistinguishable from idiopathic PD. The effects of the G2019S
mutation were explored in primary fibroblasts and SHSY5Y cells expressing wild type
or G2019S LRRK2. LRRK2 was predominantly in the cytosol and small vesicular
fraction of lymphoblasts and SHSY5Y cells with some localized to the mitochondria in
overexpressing cells. While we could detect LRRK2 in various mouse and marmoset
brain regions LRRK2 protein levels were higher in fibroblast and lymphoblast
cultures. LRRK2 cellular distribution, mRNA and protein expression were not
affected by the mutation.
Mitochondrial abnormalities are a common feature in PD. To determine
whether mitochondrial function is compromised in mutant cells, a detailed
bioenergetic assessment was carried out on both G2019S cell models. An increase in
basal and oligomycin inhibited respiration rates, reduced mitochondrial membrane
potential and cellular ATP levels was observed for G2019S fibroblasts with similar
changes observed in the neuroblastoma G2019S model. Respiratory rates and
membrane potential were restored with LRRK2 kinase inhibition. Our data is
consistent with reversable uncoupling of oxidative phosphorylation.
Investigating transcriptional levels of mitochondrial uncoupling proteins
(UCP) identified a G2019S dependent increase in UCP2 and 4 in fibroblasts and
SHSY5Y cells. Upstream of this transcriptional event, an interaction between LRRK2
and the negative regulator of PGC1α expression, HDAC5 was investigated for both
the wild type and G2019S protein.
We have identified a role for endogenous LRRK2 in regulating mitochondrial
bioenergetics with a kinase dependent gain of function for the G2019S mutation
4
consistent with partial uncoupling of oxidative phosphorylation. LRRK2 G2019S
enhanced the HDAC5 association which may be responsible for increased PGC1α
expression and the downstream UCP transcription linked with the observed
mitochondrial phenotype.
5
Acknowledgements
First, I would like to thank my supervisor Prof. Tony Schapira for giving me
the opportunity to work at the Clinical Neuroscience Department on a stimulating
project and in such a pleasant environment. Thank you for your guidance, support
and the freedom to express my thoughts and ideas. I would also like to express my
gratitude to Dr. Mark Cooper for his dedication to my work, for the time, effort,
ideas and concerns throughout the last three years. Thank you for teaching me how
to think, to write and to love what I do. I would also like to thank my third supervisor
Prof. John Hardy for integrating me into his department at the Institute of Neurology
and teaching me to appreciate genetics.
Thank you to all my colleagues at the Royal Free Hospital, in particular
Lydia Alvarez and Matthew Gegg who have shared this experience with me, for their
continuous support and ideas which helped move the project forward. Thank you to
Michael Cleeter and David Chau for their management of the department, the
technical support and rapid response to all demands and queries, I don’t know how
you do it. Thank you to the more recent lab members; Anna Migdalska, Joana
Magalhaes and Catharina Joana Casper as well as visiting staff; Dominic Piston,
Davor Ivankovic, Tanja Hering and Marine Bourlin who integrated and expanded our
wonderful science family. Thank you for every pub and staff night out, they were all
amazing!
A special thank you to the Parkinson’s disease patients and their relatives as
well as Prof. Dan Healy, Dr. Jan Willem Taanman and Dr. Susannah Horan involved in
the generation of the fibroblast and lymphoblast cultures. Thank you to Dr. Mark
Cooper and Dr. David Chau for generating the LRRK2 stable SHSY5Y cells, to Dr. Zhi
Yao and Dr. Gyorgy Szabadkai for the DJ1 knockdown SHSY5Y cells. Thank you to
Prof. Dario Alessi and Dr. Franscisco Inesta Vaquera for the LRRK2 antibodies, LRRK2
6
inhibitors, their help with the recombinant UCP2 production, the LRRK2 kinase
assays and my wonderful experience in Dundee. Thank you to Dr. Rina
Bandopadhyay for the help with the immunohistochemistry and the time spent at
the Queen Square Brain Bank. Thank you to Prof. Michael Duchen for his help with
the confocal analysis. Thank you to Prof. Dmitri Papkovsky (my dad) and his
wonderful lab in Ireland for their collaboration on the phosphorescent analysis, I
really enjoyed being home and working with you. Thank you to Janet North for her
help with the FACS and data analysis.
A special thank you to Eisai Limited who funded the work. In particular,
thank you to Dr. James Staddon and Dr. Malcolm Roberts for their continuous
support and intellectual input into the project. Thank you also for the LRRK2
antibodies, the Cellomics equipment and the wonderful time spent at Eisai in UCL. I
would also like to thank those involved in the UKDPC Consortium for giving me the
opportunity to present my work, share ideas and experience Parkinson’s research
from a different perspective.
Thank you to my parents, Dmitri and Natalia Papkovsky who have supported
and encouraged me at every stage of this experience. A special thank you to my
sister Katia Papkovskaia who has been a wonderful friend and housemate
throughout my PhD. Thank you to all my friends, the ones from my past and also the
many new ones I have made over the last three years. You have all contributed to
my wonderful memories.
7
Table of contents
List of tables……………………………………………………………………………………………………….. 17
List of figures………………………………………………………………………………………………………..18
List of Abbreviations……………………………………………………………………………………………..23
Chapter 1: Introduction
1.1.1 Basal ganglia and Parkinson’s disease……………………………………………………… 29
1.1.2 Pathology in PD……………………………………………………………………………………………30
1.2 Causative factors in PD…………………………………………………………………………………..32
1.2.1.1 Autosomal dominant PD genes: α-synuclein………………………………………….. 32
1.2.1.2 Autosomal dominant PD genes: Leucine Rich Repeat Kinase 2……………… 33
1.2.1.3 Autosomal recessive PD genes: Parkin……………………………………………………. 35
1.2.1.4 Autosomal recessive PD genes: PINK1…………………………………………………….. 35
1.2.1.5 Autosomal recessive PD genes: DJ1………………………………………………………… 36
1.2.1.6 Autosomal recessive PD genes: Glucocerebrosidase……………………………… 37
1.2.2 PD genetic models……………………………………………………………………………………… 37
1.2.3 Environmental factors in PD………………………………………………………………………..39
1.2.4 Mitochondria……………………………………………………………………………………………….39
1.2.4.1 Mechanism of action of mitochondrial toxins………………… ……………............ 42
1.2.4.2 Electron transport chain and PD……………………………………………………………… 42
1.2.4.3 Mitochondrial turnover and dynamics…………………………… ………………………. 44
1.2.4.4 Mitochondrial transcriptional regulation…………………………………................ 45
1.2.4.5 Mitochondrial and oxidative stress ………………………………………………………….46
1.2.4.6 Oxidative stress and mitochondrial uncoupling proteins………………………… 47
1.2.5 Lysosomal function and α-synuclein………………………………………………………….. 50
1.3 LRRK2……………………………………………………………………………………………………………..51
1.3.1 LRRK2 protein structure……………………………………………………………………………… 51
8
1.3.2 LRRK2 kinase ……………………………………………………………………………………………… 53
1.3.3 LRRK2 PD mutations ………………………………………………………………………………….. 54
1.3.4 LRRK2 PD mutations and neurotoxicity……………………………………………………… 55
1.3.5 LRRK2 models and PD………………………………………………………………………………… 55
1.3.6 LRRK2 detection antibodies……………………………………………………………………….. 57
1.3.7 LRRK2 mRNA and protein distribution………………………………………………………. 58
1.3.8 LRRK2 cellular distribution…………………………………………………………………………. 58
1.3.9 LRRK2 function…………………………………………………………………………………………… 59
1.3.9.1 LRRK2 and vesicles………………………………………………………………………………….. 59
1.3.9.2 LRRK2 and microtubules…………………………………………………………………………..60
1.3.9.3 LRRK2 transcription and translation ………………………………………………………. 60
1.3.9.4 LRRK2 and mitochondria ………………………………………………………………………… 62
1.4 Hypothesis …………………………………………………………………………………………………….64
1.5 Aims……………………………………………………………………………………………………………… 64
Chapter 2: Materials and Methods
2.1 Plasmid constructs………………………………………………………………………………………… 66
2.2 Brain samples……………………………………………………………………………………………….. 66
2.3 Cell lines…………………………………………………………………………………………………………67
2.4 Cell culture……………………………………………………………………………………………………. 69
2.5 Transformation of DH5a cells and DNA extraction……………………………………… 70
2.6 Concentration and purity of nucleic acids…………………………………………………….. 72
2.7 Restriction enzyme digest of pET-DEST51……………………………………………………. 72
2.8 Agarose gel electrophoresis………………………………………………………………………….. 73
2.9 Transfection of mammalian cells…………………………………………………………………. 73
2.10 Reverse transcriptase……………………………………………………………………… …………..75
2.11 Real Time PCR……………………………………………………………………………………………… 76
9
2.12 LRRK2 mutation screen in fibroblasts and SHSY5Y cells……………………………. 78
2.13 Whole cell and tissue extractions……………………………………………………………….. 78
2.14 Immunoprecipitation………………………………………………………………………………….. 79
2.15 Protein quantification…………………………………………………………………………………. 80
2.16 SDS PAGE……………………………………………………………………………………………………..80
2.17 Western blot analysis………………………………………………………………………………….. 81
2.18 Native gel electrophoresis…………………………………………………………………………… 81
2.19 Immunofluorescence of cultured cells……………………………………………………….. 82
2.20 Immunohistochemistry on human and mouse brains………………………………… 83
2.21 LRRK2 kinase activity assays……………………………………………………………………….. 84
2.22 Isolation of PMBCs and cell sorting……………………………………………………………... 85
2.23 Cellular fractionation…………………………………………………………………………………… 87
2.24 Affinity purification of mitochondria………………………………………………………….. 87
2.25 Mitochondrial sub-fractionation…………………………………………………………………. 88
2.26 Proteinase K digestion of mitochondria……………………………………………………… 89
2.27 Single cell analysis by confocal microscopy……………………………………………….. 89
2.27.1.1 Mitochondrial membrane potential: TMRM…………………………………………. 89
2.27.1.2 Method validation for TMRM……………………………………………………………….. 90
2.27.2 Mitochondrial content……………………………………………………………………………… 92
2.27.3 GFP…………………………………………………………………………………………………………… 93
2.28.1.1 Cellular respiration: oxygen electrode…………………………………………………… 94
2.28.1.2 Method validation for the oxygen electrode………………………………………… 94
2.28.2.1 Cellular respiration: phosphorescent analysis…………………………… …………..96
2.28.2.2 Method validation of phosphorescent analysis…………………………………….. 97
2.29.1 Extracellular acidification………………………………………………………………………… 99
2.29.2 Method validation for extracellular acidification measurements… …………..100
10
2.30 Cellular ATP content……………………………………………………………………………………. 103
2.31 Cellular ROS production: Array Scanner……………………………………………………… 103
Chapter 3: Results
Characterization of LRRK2 in cell models and brain tissue
3.1. Western blot analysis of LRRK2 protein expression in cell models and brain
tissue………………………………………………………………………………………………………………….. 105
3.2 Characterization of LRRK2-V5 over-expressing SHSY5Y cells………………………
108
3.2.1 Western blot analysis of LRRK2 protein in SHSY5Y cells…………………………..
108
3.2.2 LRRK2 mutation screen in SHSY5Y cells………………………………………… …………..109
3.2.3 Western blot analysis of LRRK2 sub-cellular distribution in SHSY5Y
cells…………………………………………………………………………………………………………………….. 111
3.2.4 Immunocytochemistry of LRRK2 sub-cellular distribution in SHSY5Y
cells…………………………………………………………………………………………………………………….. 113
3.3 Characterization of endogenous LRRK2 in lymphoblasts……………………………... 115
3.3.1 Western blot analysis of LRRK2 protein in lymphoblasts……………… …………..115
3.3.2
Western
blot
analysis
of
LRRK2
sub-cellular
distribution
in
lymphoblasts..…………………………………………………………………………………………………….. 118
3.4 LRRK2 immunoprecipitation…………………………………………………………………………..120
3.4.1
Characterization
of
LRRK2
immunoprecipitation
for
quantitative
comparison…………………………………………………………………………………………………………..120
3.4.2 LRRK2 kinase activity in lymphoblasts and SHSY5Y cells…………………………… 123
3.5 Characterization of fibroblast cell lines …………………………………………… …………..126
3.5.1 Immunoprecipitation of LRRK2 protein from fibroblasts…………………………… 126
3.5.2 LRRK2 mutation screen in control and G2019S fibroblasts……………………….. 127
3.5.3 LRRK2 mRNA expression in control and G2019S fibroblasts………… …………..127
3.6 Analysis of LRRK2 knockdown in SHSY5Y cells……………………………………………… 129
11
3.7 Immunoprecipitation of endogenous LRRK2 protein from brain………………….. 132
3.8 Comparison of LRRK2 expression in cell models and brain tissue…… …………..136
3.9
Analysis
of
LRRK2
protein
expression
by
blue
native
PAGE
electrophoresis…………………………………………………………………………………………………….137
3.10 Immunohistochemical analysis of LRRK2 protein in brain…………………………. 141
Chapter 4: Results
LRRK2 and mitochondrial function
4.1 Mitochondrial function in fibroblast cell lines………………………………………………. 145
4.1.1 The analysis of mitochondrial respiration in control and G2019S
fibroblasts…………………………………………………………………………………………………………… 145
4.1.2 The analysis of mitochondrial content in control and G2019S LRRK2
fibroblasts…………………………………………………………………………………………………………… 147
4.1.3 The analysis of mitochondrial membrane potential in control and G2019S
LRRK2 fibroblasts………………………………………………………………………………………………… 149
4.1.4 The analysis of cellular ATP content in control and G2019S LRRK2
fibroblasts…………………………………………………………………………………………………………… 150
4.1.5 The analysis of extracellular acidification rates in control and G2019S LRRK2
fibroblasts…………………………………………………………………………………………………………… 151
4.1.6 The analysis of cellular ROS in control and G2019S LRRK2 fibroblasts……..
152
4.1.7 The analysis of mitochondrial proton leak in control and G2019S LRRK2
fibroblasts…………………………………………………………………………………………………………… 153
4.1.8 The analysis of mitochondrial morphology in control and G2019S LRRK2
fibroblasts…………………………………………………………………………………………………………… 154
4.2 Mitochondrial dysfunction models………………………………………………………………. 156
4.2.1 The use of control fibroblasts treated with rotenone or antimycin A as a model
of respiratory chain dysfunction…………………………………………………………………………. 156
12
4.2.2 The use of control fibroblasts treated with rotenone and oligomycin to
evaluate the influence of ETC inhibition on the mitochondrial ATPase……………… 157
4.2.3 The use of control fibroblasts cultured in galactose rich conditions as a model
of limited cellular substrate supply………………………………………………………… …………. 159
4.3 The analysis of respiratory chain function in the SHSY5Y LRRK2-V5 overexpressing model………………………………………………………………………………………………… 161
4.4 LRRK2 kinase activity…………………………………………………………………………………….. 163
4.4.1 The analysis of LRRK2 kinase inhibition in the SHSY5Y LRRK2-V5 overexpressing model ………………………………………………………………………………………………..163
4.4.1.1 LRRK2-IN1……………………………………………………………………………………………….. 164
4.4.1.2 CZC25146………………………………………………………………………………………………... 164
4.4.1.3 TAE-648…………………………………………………………………………………………………… 165
4.4.2 The use of LRRK2 kinase inhibitors to evaluate effects of LRRK2 kinase
inhibition on mitochondrial function in the SHSY5Y LRRK2-V5 over-expressing
model…………………………………………………………………………………………………………………. 167
4.4.2.1 Evaluation of the effects of LRRK2 kinase inhibition on mitochondrial
respiration………………………………………………………………………………………………………….. 167
4.4.2.2 Evaluation of the effects of LRRK2 kinase inhibition on the mitochondrial
membrane potential…………………………………………………………………………………………… 167
4.5 LRRK2 mitochondrial localization……………………………………………………… …………. 169
4.5.1 Mitochondrial affinity purification……………………………………………………………… 169
4.5.1.1 Mitochondrial affinity purification versus differential centrifugation……. 171
4.5.1.2 Mitochondrial affinity purification from G2019S LRRK2-V5 over-expressing
SHSY5Y cells………………………………………………………………………………………………………… 171
4.5.1.3 Mitochondrial affinity purification, LRRK2 kinase inhibition………………….. 172
4.5.1.4 Mitochondrial affinity purification from lymphoblasts…………………………… 172
13
4.5.2.1 Sub-fractionation of mitochondrial enriched fractions from SHSY5Y LRRK2-V5
over-expressing SHSY5Y cells………………………………………………………………………………. 175
4.5.2.2 Sub-fractionation of affinity purified mitochondria from SHSY5Y LRRK2-V5
over-expressing SHSY5Y cells………………………………………………………………………………. 176
4.5.3.1 Sub-fractionation of mitochondria to determine LRRK2 membrane
association………………………………………………………………………………………………………….. 178
Chapter 5: Results
LRRK2 regulation of mitochondrial uncoupling: mechanisms
5.1.1 The analysis of SHSY5Y cells over-expressing murine GFP-UCP2……………..
182
5.1.2.1 The analysis of relative UCP mRNA expression in fibroblasts and SHSY5Y
cells.. …………………………………………………………………………………………………………………..184
5.1.2.2 The analysis of relative UCP mRNA expression in G2019S fibroblasts and
SHSY5Y cells………………………………………………………………………………………………………… 185
5.1.3 Western blot analysis of UCP2 protein in fibroblasts and SHSY5Y cells……. 186
5.1.4.1 Evaluation of the influence of genipin on UCP induced mitochondrial
depolarization in SHSY5Y cells…………………………………………………………………………….. 188
5.1.4.2 Evaluation of the influence of genipin on UCP induced mitochondrial
depolarization in fibroblasts……………………………………………………………………………….. 190
5.1.5 LRRK2 kinase activity assays with recombinant UCP2 protein……………………..192
5.1.6 The impact of UCP4 knockdown upon UCP4 mRNA levels and mitochondrial
membrane potential in SHSY5Y cells……………………………………………………… …………..194
5.1.7 The influence of LRRK2 kinase inhibition upon UCP mRNA expression in SHSY5Y
cells…………………………………………………………………………………………………………………….. 197
5.1.8 The impact of LRRK2 knockdown upon UCP4 mRNA levels and mitochondrial
membrane potential in SHSY5Y cells………………………………………………………………….. 199
5.2.1 Characterization of DJ1 knockdown SHSY5Y cells……………………………………… 200
14
5.2.2 The analysis of UCP mRNA expression in DJ1 knockdown SHSY5Y cells……
201
5.3 The influence of LRRK2 expression upon PGC1α, SOD2 and catalase mRNA in
SHSY5Y cells ……………………………………………………………………………………………………….. 203
5.4.1 The analysis of HDAC5 cellular distribution by immunohistochemistry in
SHSY5Y cells………………………………………………………………………………………………………… 204
5.4.2 The analysis of HDAC5 by Western blot in SHSY5Y cells ……………… …………. 206
5.4.3 The analysis of HDAC5 and LRRK2 interaction………………………………………….. 207
5.5.1 The analysis of DLP1 protein in SHSY5Y cells………………………………… …………..211
5.5.2 The analysis of LRRK2 and DLP1 interaction in SHSY5Y cells…………… …………..213
Chapter 6: Discussion
6.1 LRRK2 tissue distribution………………………………………………………………………………. 216
6.2 LRRK2 detection……………………………………………………………………………………………. 218
6.2.1 LRRK2 detection by Immunohistochemistry………………………………………………. 218
6.2.2 LRRK2 detection by Western blotting………………………………………………………… 219
6.2.3 LRRK2 detection by BN PAGE…………………………………………………………………….. 221
6.2.4 LRRK2 detection by immunoprecipitation and Western blotting…… …………. 223
6.3 Evaluation of LRRK2 protein expression in cell models and brain tissue……… 225
6.4 Evaluation of LRRK2 G2019S protein and mRNA expression……………………….. 228
6.5 LRRK2 cellular localization………………………………………………………………… …………..230
6.5.1 Wild type and G2019S LRRK2 cellular distribution……………………………………. 230
6.5.2 Wild type and G2019S LRRK2 mitochondrial localization………………………….. 232
6.5.3 LRRK2 localization in the mitochondria……………………………………………………… 233
6.6 Wild type and G2019S LRRK2 kinase activity……………………………………………….. 234
6.7 LRRK2 and mitochondrial function………………………………………………………………… 236
6.8 LRRK2 and mitochondrial permeability…………………………………………………………. 239
6.9 LRRK2 regulation of mitochondrial uncoupling proteins………………………………. 241
15
6.9.1.1 LRRK2 regulation of UCP mRNA expression………………………………… …………. 241
6.9.1.2 LRRK2 regulation of UCP protein…………………………………………………………….. 242
6.9.1.3 LRRK2 regulation of PGC1α expression ……………………………………… …………..244
6.9.1.4 LRRK2 regulation of HDAC5 ……………………………………………………………………. 246
6.9.1.5 LRRK2 GTPase mutations and mitochondrial function………………… …………..250
6.9.2 LRRK2 regulation of DJ1 …………………………………………………………………………….. 252
6.9.3 LRRK2 regulation of NFκβ …………………………………………………………………………. 254
6.9.4 UCP4 knockdown and LRRK2 regulation of Argonaute 2………………………….. 254
6.9.5 LRRK2 regulation of UCP phosphorylation………………………………………………….. 256
6.10 LRRK2 regulation of mitochondrial morphology…………………………………………. 256
6.11 LRRK2, mitochondrial uncoupling and cellular disturbances implicated in
PD………………………………………………………………………………………………………………………..260
6.11.1 LRRK2 regulation of ATP, mitochondrial biogenesis and calcium…………… 260
6.11.2 LRRK2 regulation of autophagy and mitophagy………………………………………. 261
6.11.3 LRRK2 regulation of synaptic function and tubulin dynamics………………….. 263
6.12 Mitochondrial uncoupling proteins and disease………………………………………….. 264
6.13 Endogenous LRRK2 regulation of UCPs………………………..……………………………… 266
List of references references ……………………..………………………………………………………. 268
Appendix ……………………..……………………………………………………..………………………………301
16
List of Tables
Table 2.1 Brain sample information describing the available tissue regions, source
and post mortem (PM) delay of samples.
Table 2.2 Information relating to the sources of punch skin biopsies and peripheral
blood mononuclear cells from which the fibroblast and lymphoblast cell lines were
derived.
Table A1 List of siRNA and shRNA sequences used in siRNA knockdown experiments.
Table A2 List of primers used for PCR analyses.
Table A3 List of antibodies
17
List of Figures
Figure 1.1 Schematic summary of the basal ganglia model.
Figure 1.2 Schematic summary of the Braak model of pathological spread in PD.
Figure 1.3 Schematic representation of the mitochondrial respiratory chain.
Figure 1.4 Schematic diagram of the LRRK2 structure.
Figure 2.1 Optimization of siRNA delivery into SHSY5Y cells by the method of
transient transfection.
Figure 2.2 Analysis of gene expression by quantitative real time PCR.
Figure 2.3 Analysis of the mitochondrial membrane potential by TMRM staining.
Figure 2.4 Analysis of mitochondrial content by TMRM staining.
Figure 2.5 Analysis of cellular respiration by the Clark type oxygen electrode.
Figure 2.6 Analysis of cellular respiration by phosphorescent oxygen probes.
Figure 2.7 Analysis of cellular acidification by phosphorescent pH sensitive probes.
Figure 2.8 The analysis of ROS by measuring the rate of dihydroethidium oxidation.
Figure 3.1 LRRK2 antibody screen.
Figure 3.2 Characterization of SHSY5Y cells over-expressing LRRK2.
Figure 3.3 LRRK2 sub-cellular distribution analysed by differential centrifugation and
Western blot analysis.
Figure 3.4 LRRK2 sub-cellular distribution by immunocytochemistry in SHSY5Y cells.
Figure 3.5 Expression of endogenous LRRK2 protein in EBV transformed
lymphoblasts and lymphocytes.
Figure 3.6 LRRK2 sub-cellular distribution in control lymphoblasts.
Figure 3.7 Validation of LRRK2 immunoprecipitation (IP) for use in quantitative
analysis of protein levels.
Figure 3.8 Analysis of LRRK2 kinase activity in EBV transformed lymphoblasts and
SHSY5Y cells.
18
Figure 3.9 Characterization of control and G2019S LRRK2 fibroblasts.
Figure 3.10 Analysis of LRRK2 knockdown in SHSY5Y cells.
Figure
3.11
Analysis
of
LRRK2
expression
in
the
human
brain
by
immunoprecipitation.
Figure 3.12 Analysis of LRRK2 expression in rodent and non human primate brains by
immunoprecipitation.
Figure 3.13 Comparison of LRRK2 expression in cell models and brain tissue.
Figure 3.14 Analysis of LRRK2 protein expression by blue native PAGE
electrophoresis.
Figure 3.15 Analysis of LRRK2 protein expression in human brain by blue native page
electrophoresis.
Figure 3.16 Immunohistochemical analysis of LRRK2 protein in mouse brain.
Figure 3.17 Immunohistochemical analysis of LRRK2 protein in human brain.
Figure 4.1 The analysis of mitochondrial respiration in control and G2019S LRRK2
fibroblasts.
Figure 4.2 The analysis of mitochondrial content in control and G2019S LRRK2
fibroblasts.
Figure 4.3 Use of TMRM to measure the mitochondrial membrane potential in
control and G2019S LRRK2 fibroblasts.
Figure 4.4 The analysis of cellular ATP content in control and G2019S LRRK2
fibroblasts.
Figure 4.5 The analysis of extracellular acidification rates in control and G2019S
LRRK2 fibroblasts.
Figure 4.6 The analysis of cellular ROS in control and G2019S LRRK2 fibroblasts.
Figure 4.7 The use of oligomycin to measure the mitochondrial proton leak in
control and G2019S LRRK2 fibroblasts.
19
Figure 4.8 The use of TMRM to analyse mitochondrial morphology in control and
G2019S LRRK2 fibroblasts.
Figure 4.9 The use of control fibroblasts treated with rotenone or antimycin A as a
model of respiratory chain dysfunction.
Figure 4.10 The use of control fibroblasts treated with rotenone and oligomycin to
evaluate the influence of ETC inhibition on the mitochondrial ATPase.
Figure 4.11 The use of control fibroblasts cultured in galactose rich conditions as a
model of limited cellular substrate supply.
Figure 4.12 The analysis of respiratory chain function in the SHSY5Y LRRK2-V5 overexpressing model.
Figure 4.13 The analysis of LRRK2 kinase inhibition in the SHSY5Y LRRK2-V5 overexpressing model.
Figure 4.14 The use of LRRK2 kinase inhibitors to evaluate effects of LRRK2 kinase
inhibition on mitochondrial function in the SHSY5Y LRRK2-V5 over-expressing model.
Figure 4.15 Affinity purification of mitochondria from LRRK2-V5 over-expressing
SHSY5Y cells and lymphoblasts.
Figure 4.16 Mitochondrial sub-fractionation to identify LRRK2 mitochondrial
localization.
Figure 5.1 SHSY5Y cells over-expressing murine GFP-UCP2.
Figure 5.2 Analysis of relative UCP mRNA expression fibroblasts and SHSY5Y cells.
Figure 5.3 Expression of UCP mRNA in G2019S fibroblasts and SHSY5Y cells.
Figure 5.4 Western blot analysis of UCP2 protein expression in fibroblasts and
SHSY5Y cells.
Figure 5.5 Evaluation of the influence of genipin on UCP induced mitochondrial
depolarization in SHSY5Y cells.
20
Figure 5.6 Evaluation of the influence of genipin on UCP induced mitochondrial
depolarization in fibroblasts.
Figure 5.7 The analysis of UCP2 as a LRRK2 kinase target.
Figure 5.8 The analysis of UCP4 knockdown in SHSY5Y cells using UCP4 siRNA.
Figure 5.9 Influence of UCP4 levels on the mitochondrial membrane potential in
SHSY5Y cells.
Figure 5.10 The influence of LRRK2 kinase inhibition upon UCP mRNA expression in
SHSY5Y cells.
Figure 5.11 The impact of LRRK2 knockdown upon UCP4 mRNA levels and
mitochondrial membrane potential in SHSY5Y cells.
Figure 5.12 Characterization of DJ1 knockdown SHSY5Y cells.
Figure 5.13 The analysis of UCP mRNA expression in DJ1 knockdown SHSY5Y cells.
Figure 5.14 The influence of LRRK2 expression upon PGC1α, SOD2 and catalase
expression in SHSY5Y cells.
Figure 5.15 The analysis of HDAC5 cellular distribution by immunohistochemistry in
SHSY5Y cells.
Figure 5.16 The analysis of HDAC5 cellular levels by Western blot in SHSY5Y cells.
Figure 5.17 The analysis of HDAC5 and LRRK2 interaction.
Figure 5.18 The analysis of DLP1 protein in SHSY5Y cells.
Figure 5.19 The analysis of LRRK2 and DLP1 interaction in SHSY5Y cells.
Figure A1 Chemical structures of LRRK2 kinase inhibitors
Figure A2 Agarose gel electrophoresis of RT PCR products.
Figure A3 Analysis of LRRK2 expression in fibroblasts by immunoprecipitation.
Figure A4 Agarose gel electrophoresis visualizing PCR products of the amplified
LRRK2 exon 41 gene.
21
22
List of Abbreviations
4E-BP - eIF4E-binding protein
Ψm – mitochondrial membrane potential
AB – anode buffer
AD – Alzheimer’s disease
Ago2 – Argonate 2
AmpR – ampicillin resistance
ANT - adenine nucleotide translocator
ARE - antioxidant responsive element
ATP – adenosine triphosphate
ATP13A2 - ATPase type 13A2
AV – autophagic vacuole
BCA - b
BN PAGE – blue native polyacrylamide gel electrophoresis
BSA – bovine serum albumin
CAMKK2 – calmodulin like kinase kinase 2
CB – cathode buffer
cDNA – coding DNA
CHEK2 - checkpoint kinase 2
CMA – chaperone mediated autophagy
CMV - cytomegalovirus
COR - C-terminal of ROC
CT – cycle threshhold
DA - dopamine
DAB - 3,3' -diaminobenzidine
DAPI - 4',6-diamidino-2-phenylindole
23
DAT – dopamine transporter
DDM - n-dodecyl-β-D-maltopyranoside
DHE - dihidroethidium
DMSO - dimethylsulfoxide
DMEM - Dulbecco's Modified Eagle Medium
DLB – dementia with Lewy bodies
dLRRK2 – Drosophila melanogaster homologue of LRRK2
DLP1 - dynamin like protein 1
dNTP - deoxyribonucleotide triphosphate
DTT - dithiothreitol
EBV - Epstein-Barr Virus
EDTA - ethylenediaminetetraacetic acid
EF-1α – elongation factor 1 alpha
EF2α – elongation factor 2 alpha
ER – endoplasmic reticulum
FADD - Fas-Associated protein with Death Domain
FADH2 - flavin adenine dinucleotide hydroquinone
FCCP - trifluorocarbonylcyanide Phenylhydrazone
Fis1 - mitochondrial fission 1 protein
FSD – full scale deflection
GAK - Cyclin G-associated kinase
GCase - glucocerebrosidase
GSH - glutathione
GSK3β - Glycogen synthase kinase 3 beta
GST - LRRK2- 1326 to 2527 residues of the LRRK2 sequence conjugated to GST
GTP - guanosine-5'-triphosphate
24
[H+] - proton
H2O2 - hydrogen peroxide
HDAC5 – histone deacetylase 5
IFNγ – interferon gamma
IP – immunoprecipitation
IPD – idiopathic Parkinson’s disease
Keap1 - kelch-like ECH-associated protein 1
KLH sequence - keyhole limpet hemocyanin sequence
KmR – kanamycin resistance
LPS - lipopolysaccharide
Lrk-1 – Carnobaelis elegans homologue of LRRK2
LRRK1 – Leucine Rich Repeat Kinase 1
LRRK2 - Leucine Rich Repeat Kinase 2
MAOB – monoamine oxidase B
MAPK7 – mitogen activated kinase 7
MAPKKK – mitogen activated kinase kinase kinase
MBP – myosin basic protein
MEF – mouse embryonic fibroblasts
MEF2 - myocyte enhancer factor-2
MeOH - methanol
Mfn - mitofusin
MVB – multivesicular bodies
MYLK - myosin light chain kinase
NAD(P)H - nicotinamide adenine dinucleotide (phosphate)
NFAT - nuclear factor of activated T cells
NFκB - nuclear factor kappa B
25
NGS - normal goat serum
NMDA - N-methyl-D-aspartic acid
NO – nitric oxide
NP-40 - nonidet P40
Nrf2 - nuclear erythroid related factor-2
NRON - repressor of the nuclear factor of activated T cells
O2.- - superoxide anion
OCT – optimal cutting temperature
OH.- - hydroxyl radical
OONO- - peroxynitrate
Opa1 - optic atrophy type 1
PBS – phosphate buffer saline
PBS-T – phosphate buffer saline supplemented with Tween
PCR – polymerase chain reaction
PD – Parkinson’s disease
PFA – paraformaldehyde
PGC1α - peroxisome proliferator-activated receptor-γ coactivator 1-α
PINK1 - PTEN-induced putative kinase 1
PKC – protein kinase C
PKA – protein kinase A
PMA - phorbol 12-myristate 13-acetate
PMBC – peripheral blood mononuclear cells
PMSF - phenylmethanesulfonylfluoride
PNS – post nuclear supernatant
PLK4 - polo-like kinase 4
26
PPAR - peroxisome proliferator-activated receptor
PSF - polypyrimidine tract-binding protein associated splicing factor
PTEN - phosphatase and tensin homolog
RFU – relative fluorescent units
RISC - RNA-induced silencing complex
ROC - Ras of complex proteins
ROS – reactive oxygen species
RPU – relative phosphorscent units
RPMI - Roswell park memorial institute medium
RT PCR – real time PCR
RXR - retinoid X receptor
SA - specific activity
SDHA - succinate dehydrogenase complex, subunit A
SDS - sodium dodecyl sulphate
SDS PAGE - sodium dodecyl sulfate polyacrylamide gel electrophoresis
SNARE - soluble N-ethylmaleimide sensitive factor attachment protein receptor
SNpc – substantia nigra pars compacta
SOD2 – superoxide dismutase 2
TFAM – mitochondrial transcription factor
TMRM - tetramethyl rhodamine methyl ester
TR-F – time resolved fluorescence
27
UCP – uncoupling protein
VDAC - voltage-dependent anion channels
28
Chapter 1: Introduction
1.1.1 Basal Ganglia and Parkinson’s Disease
Parkinson’s disease (PD) is the second most common neurological disorder
affecting approximately 2% of the population over the age of 60 and 4% over 80
years [1]. Clinical motor PD abnormalities are thought to result predominantly from
the loss of dopaminergic neurons in the substantia nigra pars compacta (SNpc).
Dopamine (DA) is a catecholamine, and the primary function of this modulatory
neurotransmitter is to serve as the reinforcement signal for the learning and
maintenance of adaptive behaviours. Introducing a stimulus to the striatum, nigral
dopamine regulates striatal output to the thalamus through the intricate circuitry of
the basal ganglia [2] (Fig 1.1, adapted from [3]).The pathophysiology and resulting
clinical manifestation in PD is thought to arise as a result of striatal dopamine
depletion, increased neuronal activity in the output nuclei of the basal ganglia and
inhibition of thalamocortical responses. Suppression of thalamic neurons is required
to maintain control of unwanted physical movements. Typical symptoms of
hypokinesia, akinesia and bradykinesia develop due to this shift in basal ganglia
function towards inhibitory cortically aided movements. The resulting loss of
regulatory control leads to deterioration of simultaneous and repetitive movements
with a decrease in amplitude of repetitions [4]. Tremor and rigidity seem to result
from altered oscillatory activity in the basal ganglia circuitry, due to
desynchronization of positive and negative regulatory pathways [5]. At the time of
clinical manifestation it has been estimated that about 50-80 % of nigral
dopaminergic neurons have been lost potentially accounting for the motor deficit.
As well as regulation of motor function, the basal ganglia have been
proposed to have a role in cognitive control [6], potentially explaining the cognitive
malfunction in later stages of PD, although this could also be due to spread of PD
29
pathology to cortical areas (section 1.1.2). Cognitive symptoms are linked to the
deterioration of cortical neurons in progressive disease stages. As well as motor and
cognitive defects, affected subjects often develop autonomic disturbances.
Gastrointestinal, cardiovascular and urogenital systems, under the control of
cholinergic, noradrenergic and serotonergic nuclei can suffer, emphasising the
contribution of other cell populations to the disease [5].
1.1.2 Pathology in PD
The pathological hallmarks of Parkinson’s disease are Lewy bodies and
dystrophic neurites. Lewy bodies are proteinaceous, cytoplasmic inclusion
composed
primarily
of
aggregated
α-synuclein
protein
and
positively
immunoreactive for ubiquitin [7] while dystrophic neurites appear as damaged
projections of affected cells. According to the Braak hypothesis of PD transmission,
the progressive pathology in PD is thought to occur as a result of pathological spread
30
from the brain stem to cortical regions potentially accounting for the motor deficit
and delayed cognitive decline. This transmission has been subdivided into six stages,
characterized by the appearance and spread of the neuropathological hallmarks (Fig
1.2, adapted from [8]). Pathology is initially confined to the olfactory bulb and lower
brainstem regions, spreading up to the locus coeruleus in stage II of PD. Clinical
symptoms are thought to occur in stage III where Lewy neurites, α-synuclein
aggregates and Lewy bodies can be detected in SNpc. Stage IV comes with nigral cell
loss and spread to the hippocampal and basal forebrain areas. Neocortical Lewy
body pathology occurs in stage V, confined primarily to the temporal regions and
continues to spread in stage VI of the disease [5].
Pathological diagnosis in PD requires depigmentation of the substantia nigra
as a result of neuronal cell loss and presence of Lewy bodies in post mortem brain
tissue to accompany the clinical motor abnormalities [9]. However, ‘parkinsonian’
phenotypes are not solely restricted to PD with features of idiopathic parkinsonism
(IPD) found in a wider spectrum of movement disorders [10]. Furthermore, a clinical
diagnosis of PD does not always correlate with the pathological requirements. In
particular, juvenile onset recessive forms of the disease (section 1.2) present with
more atypical clinical features that closely resemble dystonia [11]. In these cases,
disease progression is slow and pathology variable. Autosomal dominant PD (section
1.2.) can differ in pathology even between family members carrying the same
genetic mutation with Lewy body positive cases reported as well as those with tau,
β-amyloid deposits (hallmark of Alzheimer’s disease (AD)) and ubiquitin positive
inclusions [12]. Variable neuropathology in PD indicates the potential involvement of
multiple cellular pathways in the disease pathogenesis which may or may not
converge to produce a common clinical phenotype.
31
1.2 Causative factors in PD
The etiology of the disease is unknown. Idiopathic forms of PD are thought
to result from variable contributions of genetic and environmental origin, with rare
cases caused solely by one or the other factors. Gene linkage analyses and positional
cloning have identified specific mutations which confer a familial link in
approximately 10 % of all PD cases [13]. The genes identified can be subdivided into
two groups based on their inheritance pattern; autosomal dominant and autosomal
recessive.
1.2.1.1 Autosomal dominant PD genes: α-synuclein
The genetic link between α-synuclein and Parkinson’s disease was first
identified in 1996 for a large Greek/Italian kindred with a mean age of onset of 55
years [14]. A year later, the G209C mutation (A53T) was described in the α-synuclein
32
gene located on chromosome 4 for this cohort [15]. Additional familial linkages in
the α-synuclein gene were found as point mutations at positions 88 (G to C
substitution, A30P mutation) in a German kindred and position 188 in a Spanish,
Basque family (G to A substitution, E46K mutation) [16]. In addition, families carrying
duplications (French cohort) [17] and triplications (Iowa cohort) [18] of the αsynuclein gene have been reported suggesting that cellular levels of the protein
correlate with PD onset, progression and severity. Furthermore, genome-wide
association studies generated from large cohorts of idiopathic PD patients suggest
that α-synuclein sequence polymorphisms are a risk factor for the disease, in
support of the gene linkage data [19-21].
Inherited α-synuclein mutations result in similar clinical phenotypes of Ldopa responsive parkinsonism but appear more aggressive when compared to
idiopathic cases. Clinical features and pathology of point mutation and duplication
α-synuclein cases fall into the spectrum of PD variations, clinically diagnosed as
dementia with Lewy bodies (DLB) due to their progressive cognitive deficits [16, 22].
Lewy body pathology was identified for each familial α-synuclein case.
α-synuclein is an intrinsically disordered 14 kDa protein highly abundant in
the brain [23, 24]. Structurally, α-synuclein is capable of conformational changes
which govern its membrane binding properties and aggregation propensity [25, 26].
1.2.1.2 Autosomal dominant PD genes: Leucine Rich Repeat Kinase 2
Sequence analysis of large Japanese families identified a heritable
component in the PARK8 locus on chromosome 12 (12p11.2-q13.1) [27].
Subsequently, Leucine Rich Repeat Kinase 2 (LRRK2) pathogenic mutations within
linked families were mapped for North American (R1441C) and European (Y1699C)
kindreds as well as a small German family (I2020T) [28]. In parallel, families and
sporadic PD cases from the Basque country in Spain were identified to harbour
33
LRRK2 R1441G mutations while British families were linked to Y1699C substitutions
by the same group [29].
Subsequent identification of the G2019S mutation spurred interest in the PD
research field due to its high frequency in specific populations and association with
the ‘classical’, late onset sporadic PD. First reports of G2019S came from two
unrelated Caucasian families of North American-British and Ashkenazi Jewish
ancestry [30]. Reports of 13 individuals from a variety of ethnic backgrounds (North
America, Norway, Ireland and Poland) sharing a common disease associated
phenotype, confirmed G2019S occurence in sporadic PD [31]. Population studies
determined greatest G2019S prevalence in the Ashkenazi Jew and North African
Arabs, identified in 27.9 %, and 30 % of all PD cases from these regions respectively
[32, 33] with lower incidence in North American (7 % [34]) and European (6.6 % [35])
cohorts. In control groups, the highest frequency of G2019S was reported in Israel
(2.4 %) and Arab Berber Tunisians (2 %) [32, 36]. Haplotype analysis of
demographically distinct populations identified a common ancestrial founder [37].
More recently LRRK2 has been implicated in the disease through genome-wide
association studies of idiopathic PD patients in the UK and Japanese cohorts [19, 20].
In humans and rodents a paralogous gene to LRRK2 has been identified
sharing 26 % sequence identity termed Leucine Rich Repeat Kinase 1 [38]. However,
PD associated mutations in the LRRK1 gene have not been described.
The number of reported autopsies for all the mutant LRRK2 PD cases totals
38 [12]. Lewy bodies and dopaminergic cell loss have been identified in association
with all established pathogenic mutations. Lewy bodies are generally described in
the brain stem, cortex and the limbic system. A number of studies reported pure
nigral degeneration free from Lewy bodies associated with LRRK2 [39] however,
these cases are rare. Additional features include neurofibrillary tangles with the
34
degree of plaques and tangles in G2019S cases ranging from levels compatible with
normal ageing to densities observed in demented Alzheimer’s disease patients [40].
Clearcut tauopathies have been noted for some R1441C and G2019S cases [12].
Although heterogeneity can be observed in mutant LRRK2 PD, in general,
pathological hallmarks are comparable to those found in idiopathic disease [41].
1.2.1.3 Autosomal recessive PD genes: Parkin
Japanese families found to contain deletions in the PARK2 gene on
chromosome 6, led to the discovery of parkin as a causative risk gene for autosomal
recessive juvenile PD [42]. Loss of parkin protein function due to autosomal
recessive familial gene deletions have been found in European families however,
missense mutations, resulting in partial or complete loss of function, are more
common among the Caucasian haplotypes [43]. Single allele mutations, inherited in
an autosomal dominant manner have also been identified and seem to infer a risk
factor for the disease [44].
Unlike sporadic PD, PD cases associated with loss of parkin function have an
early onset and slow disease progression with good responses to dopamine
replacement therapy [42-44]. Furthermore autopsies of parkin PD patients rarely
show evidence of Lewy body pathology suggesting a distinctive cause for this
subtype of PD [45].
The 50 kDa protein encoded by the parkin gene is a ubiquitin protein ligase
with a RING finger domain [42].
1.2.1.4 Autosomal recessive PD genes: PINK1
First identified in three consanguineous Sicilian families, loss of function
mutations in the Phosphatase and Tensin (PTEN) induced kinase (PINK1) gene
located on chromosome 1 were shown to cause autosomal recessively inherited PD
[46]. Subsequently, the PINK1 homozygous L347P and a compound heterozygous
35
G240L/L489P were identified in Filipino families [47]. Mutations in PINK1 are
infrequent when compared to familial parkin cases. Similar to parkin, heterozygous
PINK1 mutations have been identified as a risk factor for PD [48]. In addition, PINK1
mutations have been found in sporadic PD subjects [49].
Clinical features of PINK1 PD present with early onset parkinsonism with
slow disease progression, good levodopa response and bradykinetic symptoms
mimicking that of parkin PD. Although only a few cases have been presented at
autopsy, PD patients with PINK1 mutations have been described with Lewy body
deposits [50].
Intracellular proteolytic processing of full length 63 kDa PINK1 generates
two additional forms of the protein; 54 kDa and 10 kDa respectively [51]. The
enzymatic domain in PINK1 has putative kinase characteristics although
physiological phosphorylation substrates have not yet been identified.
1.2.1.5 Autosomal recessive PD genes: DJ1
PD associated mutations in DJ1 were first identified in the PARK7 gene,
located on chromosome 1 in two consanguineous families from Italy and the
Netherlands. Patients from the Dutch family carried a gene deletion while Italian
subjects encoded a single base substitution (L166P) resulting in a loss of function
protein [52]. DJ1 mutations are rare in PD, however, the clinical phenotype of DJ1
patients resemble those found in other early onset, recessively linked PD cases [45].
No neuropathology in PARK7 cases has been reported to date.
The 20 kDa DJ1 protein exists as a dimer formed by a disulphide bond
between two monomeric molecules. The potential for free radical oxidation of DJ1
cysteine residues labelled the protein as a redox sensing molecule [53].
36
1.2.1.6 Autosomal recessive PD genes: Glucocerebrosidase
The glucocerebrosidase (GCase) gene located on 1q21-22, spanning 11
exons, has been reported to have >300 mutations associated with Type 1-3
Gaucher’s disease, a lysosomal storage disorder. In addition, heterozygous
mutations within the GCase are the most common risk factor for sporadic
Parkinson’s disease [54]. Mutation penetrance can vary between ethnic groups
observed at high frequency for Ashkenazi Jews [55] and low frequency in European
cohorts [56]. Initially regarded as a risk factor rather than a mendelian form,
identification of familial links has provided the criteria to label GCase as an
autosomal recessive PD gene [57, 58].
Clinically, parkinsonian symptoms in GCase PD cases occur with early onset
relative to idiopathic disease and typical Lewy body pathology has been described
for the limited number of cases inspected at autopsy [59].
The
GCase
protein,
synthesised
in
endoplasmic
reticulum
(ER)
polyribosomes is targeted to the lysosome following peptide cleavage and N linked
glycosylation. Polyubiquitination of GCase facilitates ER exit. In the lysosomal
compartment, GCase functions to break down the lipid glucoceramide into glucose
and ceramide [60]. Some of the most common GCase mutations are reported to trap
GCase in the ER depleting the lysosomal GCase pool [61].
1.2.2 PD genetic models
The discovery of familial PD genes led to the development of genetic PD
animal models in an attempt to recapitulate the clinical features and pathology of
PD and dissect the mechanism of disease pathogenesis. Although α-synuclein
transgenic mouse models have frequently been shown to develop motor
abnormalities correlating with a reduction in dopamine transmission and αsynuclein positive inclusions, these protein aggregates were reported to be distinct
37
from Lewy bodies described in PD patients [62-65]. In addition, dopaminergic
neuronal loss was not observed in the animals. Defects in mitochondrial morphology
and mitochondrial dysfunction have been described in two studies [66, 67]
suggesting a potential role of mitochondria in the α-synuclein pathology and motor
disturbances associated with α-synuclein over-expression in mice. Although
C.elegans and D.melanogaster lack α-synuclein orthologues, over-expression of the
human wild type and mutant α-synuclein protein in flies has shown to result in age
dependent motor abnormalities, dopamine depletion and the formation of αsynuclein positive aggregates [68, 69]. Slowness of movement has been described in
C.elegans over-expressing wild type and mutant α-synuclein consistent with a
perturbation in dopaminergic function [70, 71] and correlating with the loss of
dopaminergic neurons [71] however, α-synuclein aggregates were not reported.
Recessive PD has been modelled in animals through deletions of DJ1, PINK1
and parkin genes. Although PINK1, parkin and DJ1 knockout mice have signs of
abnormal dopamine metabolism, nigral dopaminergic cell death is not a feature and
Lewy bodies have not been observed in these animals [72-74]. Drosophila
melanogaster models of loss of PINK1 and parkin function have revealed motor
defects and dramatic defects in mitochondrial morphology in energy intensive
tissues [75] as well as degeneration of dopaminergic neurons [76]. Although DJ1
knockout flies also show mitochondrial structural alterations, these appear to be
distinct from those identified in PINK1 and parkin knockouts [77]. Mild
mitochondrial abnormalities have been observed in PINK1 and parkin knockout mice
however these features are less pronounced than the corresponding fly models [78].
Although none of the animal models recapitulate the full pathological
spectrum of PD, some overlapping features indicate the involvement of distinctive
cellular pathways in the disease pathogenesis. In particular, the pathways involved
38
in α-synuclein synthesis and degradation mediating steady state levels of the
protein, synaptic function regulating dopamine metabolism and mitochondrial
function contributing to cellular energy homeostasis.
1.2.3 Environmental factors in PD
Familial mutations account for about 10 % of all PD cases suggesting
additional factors may be involved in the disease pathogenesis. Population based
studies have shown that head trauma may accelerate PD onset [79] while meta
analysis [80] of studies in humans and rodents [81] have demonstrated a protective
role for cigarette smoking and neuroprotection from nicotine. Studies looking at
physical activity, caffeine use and non steroidal anti inflammatory drugs have all
shown negative correlations with respect to the disease [82-84]. In addition,
exposure to environmental toxins and pesticides has been shown to correlate with
sporadic forms of PD [85, 86]. 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)
was identified to induce Parkinsonian like symptoms and pathology in humans and
non-human primates [85, 87]. Chronic, systemic administration of the pesticide
rotenone in rodents led to the development of PD like symptoms and α-synuclein
aggregates [88]. Paraquat, a quaternary ammonium herbicide, structurally
resembles MPTP and resulted in selective loss of dopaminergic neurons, dopamine
depletion and α-synuclein aggregation in rodents [89, 90]. The subsequent
characterization of the molecular mechanisms associated with these compounds
further emphasised the role of mitochondrial function in PD pathogenesis. The link
between mitochondrial dysfunction and α-synuclein aggregation in toxic animal and
human PD models was also noted.
1.2.4 Mitochondria
Mitochondria are double membrane bound organelles consisting of a
relatively permeable, cholesterol rich outer membrane and an impermeable, protein
39
rich inner membrane. The outer membrane houses different families of proteins
facilitating mitochondrial import, organelle dynamics and apoptotic cascades. The
cristae folds of the inner membrane allow for high concentrations of proteins
involved in processes such as oxidative phosphorylation to meet the metabolic
demands of the cell. The mitochondrial matrix incorporates mitochondrial
transcriptional machinery working in concert with nuclear genes to regulate
mitochondrial function. In addition, the mitochondrial matrix facilitates the
tricarboxylic acid cycle (TCA) supplying reducing equivalents to the electron
transport chain (ETC).
In terms of mitochondrial involvement in PD pathogenesis, a considerable
amount of attention has been focused on the mitochondrial respiratory chain.
Structurally, the mitochondrial respiratory chain consists of 5 multi subunit
complexes where Complexes I, III, IV and V span the inner mitochondrial membrane
while Complex II contacts the membrane at the matrix side (Fig 1.3, adapted from
[91]). Reducing equivalents (NAD(P)H and FADH2) from the TCA and β-oxidation
function as electron donors feeding into Complexes I and II. Redox cofactors bound
to their respective enzymes transfer electrons to the mobile electron carrier
coenzyme Q 10 which in turn is oxidised by Complex III to allow for electron transfer
to cytochrome C. Shuttling of reduced cytochrome C to Complex IV facilitates the
reduction of oxygen by this enzyme. Electron transfer by Complexes I, III and IV
results in the movement of protons [H+] across the inner membrane to the intermembrane space. The generated proton gradient is referred to as the mitochondrial
membrane potential. The thermodynamic tendency of [H+] to flow back into the
matrix generates energy, the driving force for Complex V (ATP synthase complex).
Respiratory chain defects often lead to reduced electron flow, mirrored by a drop in
40
the mitochondrial membrane potential and a build-up of reducing equivalents, a
source of potentially damaging reactive oxygen species (ROS).
The reduction of oxygen by one electron at a time produces a range of
relatively stable intermediates. One electron reduction generates the superoxide
anion (O2.-) and spontaneous or enzyme induced dismutation of O2.- produces
hydrogen peroxide (H2O2) which in turn may be partially reduced to hydroxyl radicals
(OH.-). OH.- can be re-reduced by O2.- propagating the process [92]. Hydroxyl radicals
can directly lead to lipid peroxidation and react with nucleic acids of mitochondrial
DNA to cause DNA damage. The superoxide anion, although relatively non-toxic can
react with nitric oxide (NO.-) to generate the highly reactive peroxynitrate (OONO-),
damaging to lipids and DNA.
41
1.2.4.1 Mechanism of action of mitochondrial toxins
MPTP selectivity for dopaminergic neurons is conferred by its affinity for the
dopamine transporter (DAT) [93]. Lipophilic properties target it to the mitochondrial
matrix where MPTP is metabolized by the enzyme monoamine oxidase B (MAO-B) to
MPP+. MPP+ blocks electron transfer through Complex I of the ETC by adhering in
close proximity to the coenzyme Q 10 binding site [94]. Additional effects on TCA
cycle enzymes results in an energy crisis believed to lead to acute dopaminergic cell
death.
Rotenone is a well characterized inhibitor of mitochondrial Complex I.
Similar to MPTP, its lipophilic nature allows it to cross the blood brain barrier,
cellular membranes and accumulate in the mitochondria. Unlike MPTP, rotenone
uptake is not restricted to dopaminergic neurons however, the enhanced
vulnerability of these neurons to rotenone toxicity links mitochondrial function to
PD pathology. Rotenone induced Complex I inhibition has been reported to enhance
ROS production implicating ROS in the disease mechanism.
Paraquat uptake is not mediated by DAT hence the molecular selectivity for
dopaminergic neurons is different to that of MPTP [95]. Paraquat is believed to kill
cells through failure of antioxidant defences and oxidative damage to cytosolic
proteins rather than the primary defect in Complex I function as is the case for
rotenone and MPP+ [96], further emphasising the role of ROS in disease
pathogenesis.
1.2.4.2 Electron transport chain and PD
The link between mitochondrial toxins and PD pathology led to the
identification of mitochondrial respiratory defects in idiopathic PD patients. Complex
I functional abnormalities were identified in the substantia nigra of IPD subjects and
confirmed by multiple groups [97-100]. Altered Complex I enzyme assembly has
42
been reported in IPD patient brains [101]. Subsequently, Complex II defects were
reported as well as Complex II in combination with Complex III or Complex IV
abnormalities of mitochondria extracted from skeletal muscle, lymphocytes and
patelets of PD subjects [102-106]. Although substantial variability was reported
between groups, the association between respiratory chain defects and disease
were consistent.
Oxidative stress was another prominent feature identified in PD patient
brain tissue similar to the effects observed in rotenone and paraquat models.
Increased levels of lipid oxidation [107], oxidative protein damage [108, 109] and
oxidative DNA damage [108, 110] have been reported in the SNpc of idiopathic PD
patients. Oxidative stress has also been noted in peripheral tissues [110-114]
suggesting that increased ROS generation is not restricted to the affected brain
areas.
Respiratory chain defects have also been noted in genetic models of PD.
Primary cells from recessive PD patients as well as DJ1, PINK1 and parkin knockout
models show evidence of respiratory chain abnormalities, particularly in Complex I
activity [115-122] and increased levels of oxidative stress [118, 120, 121]. αsynuclein association with the inner mitochondrial membrane, specifically Complex I
of the ETC is believed to modulate the bioenergetics changes associated with
reduced activity of Complex I and increased levels of oxidative stress in α-synuclein
over-expressing primary neurons [123]. Furthermore, the mitochondrial association
and Complex I inhibition is thought to be exacerbated by the toxic oligomeric and
aggregated α-synuclein [124], further emphasising the link between α-synuclein
aggregation and PD pathogenesis.
43
1.2.4.3 Mitochondrial turnover and dynamics
Discovery of the functional roles of the recessively linked PD proteins, PINK1
and parkin, has led to the mechanistic insight into a novel pathway linking the
fundamental processes involved in mitochondrial homeostasis- mitochondrial
dynamics and turnover.
Mitochondria are now recognised as dynamic structures continuously
undergoing fission and fusion events. The structural re-organisation of the
mitochondrial network involves both the inner and outer mitochondrial membranes.
Mitochondrial fusion events can dilute out mitochondrial DNA mutations and
metabolic insufficiencies [125]. Outer membrane fusion events require the action of
mitofusin 1 and 2 (Mfn) proteins while the inner membrane is fused through the
action of Optic Atrophy 1 (Opa1) protein. Fission of both membranes is facilitated by
mitochondrial recruitment of the cytosolic proteins dynamin like protein 1 (DLP1)
and mitochondrial fission protein (Fis1). Oligomerization of DLP1 forms membrane
boundaries and allows for mitochondrial budding and fragmentation [126].
Mitophagy is the process by which the mitochondrial network gets
degraded. Parkin has been shown to be involved in the removal of damaged
mitochondrial structures [127]. Depolarization of the mitochondrial membrane in
response to chemical uncoupling or electron transport chain inhibitors results in
parkin targeting from its native cytosolic location to the damaged regions of the
mitochondrial outer membrane [128]. At the outer membrane, parkin promotes
ubiquitin mediated degradation of the mitofusins and is thought to induce
mitophagy by localized prevention of mitochondrial fusion events [129]. Isolated
mitochondrial fragments targeted for mitophagic removal are subsequently
incorporated into double membrane vesicles and shuttled to the lysosome for
proteolytic processing [126]. PD linked parkin mutations interfere with various
44
aspects of this mechanism, affecting parkin translocation, ubiquitin ligase activity
and mitophagic membrane formation [130].
Functionally PINK1 has been shown to act upstream of parkin in the
pathway of mitochondrial turnover. PINK1 contains a mitochondrial localization
motif importing the protein in a membrane potential dependent mechanism where
it is rapidly cleaved by mitochondrial proteases generating the truncated PINK1
forms [131]. Mitochondrial depolarization leads to the accumulation of full length
PINK1 at the outer surface, believed to act as the signal for parkin recruitment [132].
Mutations in PINK1 tend to either destabilize the protein structure or occur in the
enzymatic domain, leading to a loss of function phenotype [46, 47].
The ETC and mitochondrial morphology defects described for PD linked
PINK1 and parkin mutations are thought to result from insufficient removal of
damaged organelles [129, 133]. In addition, parkin is believed to mediate GCase
polyubiquitination in the ER and facilitate GCase exit [134]. Excessive parkin
occupancy of trapped mutant GCase is believed to deplete the pool required for
mitophagy resulting in the build up of damaged organelles [134]. Altered
mitochondrial membrane binding properties of α-synuclein mutants are thought to
account for the mitochondrial fragmentation in cells [135, 136]. α-synuclein
membrane binding could thus influence mitochondrial fusion by interfering with the
mitochondrial fission fusion machinery. Hence dysregulation of mitophagy has been
linked to most of the genetic PD models and appears to contribute to PD
pathogenesis.
1.2.4.4 Mitochondrial transcriptional regulation
In addition to homeostatic maintenance of healthy mitochondria through
turnover, fission and fusion events, recessive PD proteins have been implicated in an
upstream pathway of mitochondrial control through transcriptional regulation.
45
Mitochondrial biogenesis is the process by which the mitochondrial pool is
replenished. The main signalling pathway associated with its induction is regulated
by the nuclear factor (erythroid-derived 2), kelch-like ECH-associated protein 1
(Nrf2/Keap1) complex stimulating a battery of approximately 100 genes under the
inducible control of antioxidant response element (ARE) and myocyte enhancer
element-2 (MEF2) [137]. In addition to the induction of expression of respiratory
chain components, transport molecules, proteosomal proteins and chaperones,
global regulatory factors; peroxisome proliferator-activated receptor gamma
(PPARγ), retinoid X receptor-alpha (RXRα) and PPARγ co-activator 1 alpha (PGC1α)
are induced to regulate the expression of mitochondrial lipid pathways integral to
bioenergetic metabolism [138]. Nuclear transcription is coupled to mitochondrial
RNA synthesis through the expression of a mitochondrial transcription factor (TFAM)
[139]. The overall outcome is increased mitochondrial content, cellular respiratory
capacity and improved maintenance of existing organelles.
Transcription and replication of mitochondrial DNA was found to be
enhanced by parkin over-expression [140], while down regulation of PINK1
correlated with reduced mitochondrial content [122], potentially contributing to the
bioenergetic defect. A functional association between parkin and TFAM has been
described in neuroblastoma cells, linking the catabolic and anabolic pathways of
mitochondrial homeostasis [140] further implicating this pathway in the
pathogenesis of PD.
1.2.4.5 Mitochondria and oxidative stress
Mitochondria respond to oxidative stress by promoting synthesis of
antioxidant enzymes. The antioxidant response can be stimulated as part of the Nrf2
signalling pathway, generally mediated by PGC1α regulated expression [141, 142] of
antioxidant enzymes such as superoxide dismutase 2 (SOD2), glutathione (GSH) and
46
catalase. The Nrf2/Keap1 complex has been shown to be sensitive to cellular redox
status governing PGC1α transcription [143]. As well as induction of expression by
Keap1/Nrf2, PGC1α transcription is negatively regulated by nuclear levels of histone
deacetylase 5 (HDAC5), a transcriptional regulator shuttling between the nucleus
and cytosol [144].
The involvement of this branch of the Nrf2/Keap1 pathway in PD
pathogenesis is supported by functional association of the downstream PGC1α with
the recessively linked DJ1 protein [145]. In the nucleus, DJ1 inhibits sumolation of
pyrimidine tract binding protein-associated splicing factor (PSF), which binds PGC1α
and mediates its transcriptional activity [145]. DJ1 has been shown to have an
altered redox state in response to oxidative stress and acidic conditions which may
in turn influence the protein’s cellular localization with its nuclear translocation
likely to govern the redox transcriptional response [146]. PD linked DJ1 mutations
can affect protein cellular distribution and disrupt the PSF modulation resulting in
reduced transcription of the antioxidant enzymes [147]. In addition, PGC1α levels
respond to parkin silencing and over-expression potentially as a result of the
oxidative stress associated with the loss of parkin phenotype [148, 149].
Furthermore, increased activity of PGC1α co-regulator PPARγ has been reported to
be protective against dopaminergic cell loss in PD models [150, 151]. Although
PGC1α is not regarded as a familial PD protein, polymorphisms in the gene sequence
link it to sporadic disease [152], further emphasising the role of oxidative stress and
mitochondria in PD pathogenesis.
1.2.4.6 Oxidative stress and mitochondrial uncoupling proteins
In addition to redox maintenance by antioxidant enzymes, ROS levels can be
regulated through the action of mitochondrial uncoupling proteins (UCPs). This
family is comprised of five isoforms with structural homology between family
47
members. While UCP2 is expressed ubiquitously, UCP1 is predominantly expressed
in brown fat, mediating thermogenesis and UCP4 and 5 are the neuronal isoforms
with poorly defined functional roles. UCP3 is expressed in some but not all tissues
and the specificity for its distribution has not yet been determined [153]. UCPs exist
as dimers on the inner mitochondrial membrane forming a pore which allows
protons to flow into the mitochondrial matrix. Proton flow through the UCP channel
provides an alternative route from the mitochondrial ATPase and partially dissipates
the mitochondrial membrane potential in the form of heat release [154, 155]. Mild
uncoupling of oxidative phosphorylation upregulates ETC function and limits the
amount of reducing equivalents available for ROS production.
Regulated transcriptionally [156, 157], translationally [158] and by post
translational modification [159], UCPs provide a rapid response to the demands of
the electron transport chain. The UCPs have a relatively short half life of 40 minutes
with the levels maintained through basal and inducible transcriptional regulation
[160]. In particular, transcriptional activation occurs by PGC1α mediated activation
or as part of the nuclear factor kappa-light-chain-enhancer of activated B cells
(NFκB) pathway [156, 157, 161, 162]. Hence UCP transcription is potentially linked to
the PD associated parkin and DJ1 proteins.
Mitochondrial uncoupling proteins have been shown to play a protective
role in PD models. Neuroprotective properties of UCPs have been demonstrated in
rodents treated with MPTP where UCP2 over-expression reduced dopaminergic cell
loss [163]. In addition, chronic MPTP treatment resulted in a time and dose
dependent induction of UCP2, 4 and 5 in neuronal cells [164]. Mechanistically,
superoxide generated during respiration can result in lipid peroxidation which in
turn activates UCP expression to reduce the superoxide production, effects
presumably exacerbated in MPTP models [165]. Although both neuronal and non
48
neuronal UCPs possess properties that can potentially protect cells against various
stresses, toxic effects linking UCP over-expression and cell death have been reported
[166, 167].
In addition to the protective role in PD models, a genetic link has been
reported for the neuronal UCP isoforms and the PD associated DJ1 protein. As well
as a role in induction of antioxidant enzymes DJ1 has recently been linked to a
specific function regulating the autonomous pacemaker activity of dopaminergic
neurons in the SNpc [168]. Pacemaking is unique to SNpc DA neurons and facilitates
sustained release of dopamine necessary for the proper function of target structures
in the striatum [169]. Mitochondrial regulation of this function can potentially
explain the vulnerability of this cell population in PD. Pacemaker activity is regulated
by long lasting (L) type calcium channels where cytosolic calcium oscillations come at
a metabolic cost contributed to by oxidative phosphorylation as channel opening
requires hyperpolarized membrane potentials [170]. Mitochondrial calcium uptake
is coupled to rapid depolarization and repolarization of mitochondrial membranes
assumed to be facilitated through the action of UCPs modulating the proton
gradient [168]. DJ1 knockout animals showed a reduced amplitude and frequency of
polarization events correlating with reduced expression of UCP mRNA. Furthermore
superoxides can affect mitochondrial free calcium by regulating the expression of
UCPs suggesting a potential regulatory loop involving the redox sensing DJ1 protein.
Of note, mitochondrial calcium dysregulation has been reported in PINK1 knockout
models suggesting a potential overlap between the recessive PD genes [172]. In
support of the role of calcium in PD, MPTP and rotenone are associated with
diminished mitochondrial calcium uptake and increased cytosolic free calcium in
cultured cells [173-175]. Considering the genetic association between DJ1 and PD,
49
SNpc dopaminergic cell loss and the regulatory control of mitochondrial function, it
is possible to postulate the involvement of the UCPs in PD pathogenesis.
1.2.5 Lysosomal function and α-synuclein
Familial α-synuclein gene duplications and triplications as well as Lewy body
deposits in PD brains suggest increased levels of α-synuclein protein are a key
pathological factor in PD. Evidence of the role for mitochondrial dysfunction in αsynuclein aggregation came from studies identifying α-synuclein aggregates in MPTP
and rotenone mouse models of PD [176-179].
Steady state α-synuclein levels are determined by the balance between
protein synthesis and degradation pathways. Increased α-synuclein mRNA
expression has been shown in SNpc of IPD patients [180] suggesting increased αsynuclein expression can potentially increase cellular concentrations of the protein.
In addition, lysosomal function was determined as the rate limiting step in the
clearance of protein aggregates [181] suggesting this process may contribute to
increased α-synuclein levels in cells. α-synuclein degradation involves chaperone
mediated autophagy (CMA) which requires binding of chaperones to an intrinsic
pentapeptide motif in α-synuclein and the subsequent receptor mediated
internalization of the protein into the lysosomal lumen where it is degraded by
lysosomal proteases [182]. α-synuclein point mutations inhibit CMA [182] and may
account for the build up in protein levels.
Inhibition of CMA can stimulate another branch of the lysosomal
degradation pathway, autophagy, which involves the formation of a double
membrane vesicle around the target cargo and subsequent fusion with the
lysosomal membrane [183]. Enhanced autophagy can partially compensate for the
lack of CMA mediated degradation and has been shown to reduce protein
aggregation in a number of cellular models of neurodegenerative diseases [184-
50
187]. However enhanced autophagy can also induce cell death under stress
conditions [188].
As well as mutant α-synuclein inhibition of chaperone mediated autophagy,
lysosomal dysfunction has been linked to wild type α-synuclein accumulation.
Reduced levels of a lysosomal protease, cathepsin D, correlated with increased αsynuclein accumulation [189]. Clinical features of PD have been noted in patients
with lysosomal storage disorders accompanied by α-synuclein accumulation in
neurons and glia [190-192]. In addition, genome-wide association studies linked two
lysosomal enzymes to an increased risk for developing PD, GCase [54, 193] and a
lysosomal membrane protein probable cation-transporting ATPase 13A2 (ATP13A2).
In support of the genetic studies mutant GCase has consistently correlated with αsynuclein accumulation in cell culture models potentially providing a link to the
associated Lewy body pathology observed in GCase mutant PD cases [194, 195].
ATP13A2 deficiency is associated with impaired lysosomal degradation capacity, αsynuclein accumulation and toxicity [196, 197].
Analyses of brain tissue from subjects with lysosomal storage disorders have
also identified morphologically and biochemically deficient mitochondria [198, 199]
suggesting mitochondrial function and lysosomal regulation are intrinsically linked.
As lysosomes are also the final destination for mitophagic cargo, lysosomal
abnormalitites influence mitochondrial turnover providing a potential link to PINK1
and parkin mediated mitochondrial regulation.
1.3 LRRK2
1.3.1 LRRK2 protein structure
High incidence of LRRK2 mutations, the pathological and clinical similarities
with idiopathic disease suggest that LRRK2 plays an important mechanistic role in PD
pathogenesis.
51
The 1.4 Mb LRRK2 gene consists of 51 exons encoding a protein of 2527
amino acids. Structurally the 286 kDa LRRK2 harbours multiple functional domains;
ankyrin repeats (ANK), leucine-rich repeat domain (LRR), Ras in complex proteins
domain (ROC), C-terminal of ROC region (COR), mitogen activated protein kinase
kinase kinase (MAPKKK) domain and a WD40 domain (Fig 1.4, adapted from [200]).
Based on sequence predictions the kinase domain consists of 2 structurally
homologous lobes connected through an active hinge region. The ROC domain
contains a guanosine triphosphate (GTP) binding site while the LRRK2 COR segment
is thought to be a regulator of the ROC GTPase activity [201]. In vitro kinase and
GTPase activity measurements confirmed LRRK2 as a catalytically active molecule
[202, 203]. There is evidence to suggest the two enzymatic domains are intrinsically
linked as GTP binding to the ROC domain is important for LRRK2 kinase activity
[203]. However, GTP activity of kinase deficient LRRK2 are comparable to those of
the wild type protein and GTP concentrations do not seem to affect kinase activity
[203-206] suggesting the kinase is likely the modulator rather than output of
enzymatic activity. The flanking ankyrin repeats, LRR and WD40 are domains
regulating protein-protein interactions. WD40 contribution to the regulation of
LRRK2 enzymatic activity is supported by the lack of detectable kinase activity in C
terminally truncated LRRK2 mutants [202, 208].
The quaternary structure of LRRK2 is believed to be that of a kinase inactive
oligomer. GTP binding leads to the dissociation of the complex and heterodimer
formation with a concominant activation of the kinase. The ROC-COR region of
LRRK2 has been mapped to the dimer interface and GTPase activity reported to
mediate steady state dimer levels [209, 210].
52
1.3.2 LRRK2 kinase
As well as its inherent autophosphorylation capacity [203, 204, 211], LRRK2
can in vitro phosphorylate myelin basic protein (MBP) [203, 204] and an artificial
peptide LRRKtide, based on predicted phosphorylation sites of the LRRK2 substrates
ezrin, radixin and moesin [202].
Recently, a number of ATP competitive LRRK2 kinase inhibitors have been
developed, varying in their structure (Appendix Fig A1), selectivity properties and
potencies both in in vitro and ex vivo studies. LRRK2 IN1 [212] is a LRRK2 kinase
inhibitor with IC50=13 and 6 nM for recombinant human wild type and G2019S
LRRK2 respectively. Kinase selectivity profiling identified 12 additional targets for
LRRK2 IN1 including IC50=160 nM for mitogen activated protein kinase 7 (MAPK7 or
ERK5) and IC50>1 µM for CHK2 checkpoint homolog (CHEK2), myosin light chain
kinase (MYLK) and calmodulin binding kinase kinase 2 (CAMKK2). CZC25146 [213]
has an IC50=1-5 nM and 2-7 nM for the recombinant human wild type and G2019S
LRRK2 enzymes respectively. Kinase selectivity profiling identified 5 additional target
kinases with an IC50=10 nM-2 µM which include Polo-like kinase 4 (PLK4) and cyclin G
53
associated kinase (GAK) with no common LRRK2 IN1 targets identified in this IC50
range. TAE-684 [214] has a biochemical potency of IC50=7.8 nM and 6.1 nM for
recombinant human wild type and G2019S LRRK2. TAE-684 has been described as a
potent binder of many kinases with an IC50<100 nM for CHK2 and CAMKK2 and
considerably less selective than LRRK2 IN1 and CZC25146. Identification of 14-3-3 as
a LRRK2 interacting substrate and characterization of the phospho serine residues
required for 14-3-3 binding has revealed an inherent relationship between kinase
inhibition, 14-3-3 association and serine phosphorylation [215, 216]. These
characteristics have been used to determine the effective inhibitor doses for LRRK2
in cells and animal models [213, 214, 216].
1.3.3 LRRK2 PD mutations
PD associated mutations are located in the enzymatic domains of LRRK2.
The glycine residue at position 2019 forms part of a kinase activation motif
conserved among LRRK2 kinase homologues [217]. Substitution of glycine with a
bulkier alanine residue is thought to perturb the activation loop and produce an
overactive kinase with greater affinity for ADP. In support of the structural
predictions in vitro studies of LRRK2 enzymatic activities have consistently reported
an increase in kinase activity of the protein associated with the G2019S mutation
[201-203, 218]. Overall GTPase mutations seem to modestly raise the kinase activity
of LRRK2, however, variable effects on LRRK2 kinase function have been reported
[201-203, 218].
Mutations within the ROC-COR domain seem to induce both structural and
activity associated changes. R1441G/C and Y1699C GTPase mutations have
consistently been shown to lower the GTPase activity of the molecule [219, 220]. In
addition R1441G/C but not Y1699C GTPase mutations perturb the LRRK2 dimer
structural integrity [209, 210]. Kinase domain mutants appear not to impact on the
54
structure of the dimeric species [208, 221] but G2019S LRRK2 increases GTP
hydrolysis [222].
1.3.4 LRRK2 PD mutations and neurotoxicity
In vitro studies implicate LRRK2 enzymatic changes in nigral dopaminergic
cell death of LRRK2 PD patients. Neurotoxic effects of LRRK2 have been modelled in
cell culture systems. Neuroblastoma cells, primary rat and mouse cortical cultures
expressing mutant LRRK2 often present with nuclear shrinkage, reduced viability and
enhanced sensitivity to toxic insult [203, 223, 224]. Mechanistically, over-expressed
LRRK2 has been shown to interact with the apoptotic machinery. LRRK2 association
with Apaf1, the main component of the apoptosome - a reported initiator of the cell
death cascade [225] positions the protein downstream of a cytochrome c mediated
death pathway. LRRK2 association with the death adaptor Fas associated protein
(FADD) further emphasises the role of caspases in the toxic events associated with
the mutant protein.
An additional link to PD pathology in cell culture systems is shown by the
ability of mutant LRRK2 to aggregate in HEK293 and SHSY5Y cells over-expressing
the protein [215, 223, 224]. LRRK2 aggregates are positive for ubiquitin and could
potentially represent a form of toxic species [226]. Some but not all the pathogenic
mutants are reported to enhance aggregate formation [215, 223, 224]. However,
LRRK2 aggregates are not always observed in over-expressing SHSY5Y and HEK293
cell systems [227, 228] and identification of equivalent species in animal models and
human G2019S subjects warrants further investigation.
1.3.5 LRRK2 models and PD
Some of the pathogenic hallmarks observed with mutant LRRK2 in cell
culture systems have been extended to animal models of LRRK2 toxicity. Drosophila
melanogaster and Caenorhabditis elegans contain a single homologue of the LRRK2
55
gene (dLRRK and lrk-1), with approximately 20% sequence identity to the human
LRRK2 protein. The relevant LRRK2 homologues contain a conserved leucine rich
core kinase (LRCK) sequence and the predicted conservation of LRRK2 tertiary
structure bodes in favour of conserved cellular roles. Rodent LRRK2 homologues
share 82 % sequence identity with the human LRRK2.
Over-expression of the G2019S C.elegans lrk-1 homologue results in early
larval arrest suggesting mutant LRRK2 is potentially toxic [229]. Although overexpression of human LRRK2 is less aggressive, the mutant protein increases the
sensitivity of C. elegans dopaminergic neurons to mitochondrial toxins, suggesting
the DA neurons are less viable [230]. Motor abnormalities, alterations in brain
dopamine levels and dopaminergic cell defects are pronounced in Drosophila LRRK2
PD models [231, 232]. Defective dopamine transmission is consistently observed in
mutant human LRRK2 transgenic (G2019S and R1441G/C) rodents correlating with
an age dependent decrease in locomotor activity, responsive to L-Dopa treatment
[233-235]. Mitochondrial abnormalities were noted in LRRK2 transgenic mouse
models, associated with the over-expression of wild type and G2019S protein [233,
236]. Degeneration of axonal projections in rodent SNpc dopaminergic neurons is
similar to the axonal demise observed in cell culture and C.elegans models. These
axonal defects are thought to account for the striatal dopamine depletion and
potentially reflect the human pathological scenario [233]. The severity of the mutant
phenotype is dose dependent as mice expressing higher levels of the mutant LRRK2
protein present with more pronounced PD pathology [233]. Dopaminergic cell loss is
rarely observed and α-synuclein deposits have never been reported in any LRRK2
models.
Locomotor dysfunction, dopamine deficiency, cell toxicity and mitochondrial
abnormalities observed in LRRK2 transgenic animal models are comparable to those
56
observed for other genetic PD models, suggesting overlapping pathways in PD
pathogenesis.
1.3.6 LRRK2 detection antibodies
Although low levels of ectopic wild type LRRK2 expression do not seem to
affect cell viability, high expression has been linked to toxicity in cell culture systems,
albeit milder relative to G2019S LRRK2 [227].
When this project was initiated, LRRK2 protein characterization in the brain,
peripheral tissues and cell culture models relied on a library of in house and
commercial antibodies raised against the LRRK2 sequence. Both polyclonal and
monoclonal antibodies have been generated, targeting C and N terminal regions as
well as the core domains of the protein. Early reports utilized over-expressing
systems for antibody characterization struggling to detect low levels of endogenous
LRRK2 [237]. Subsequent validation of commercial antibodies with LRRK2 knockout
tissue identified non specific species of comparable molecular mass to full length
LRRK2 protein [237]. Additional lower molecular weight immunoreactivity is
common in LRRK2 detection by SDS PAGE, however, a limited number of these
immunoreactive bands have been characterized [238, 239].
To address the problem of endogenous LRRK2 detection, enzyme
purification for LRRK2 activity measurements and facilitate studies of LRRK2
function, Alessi et al have reported an immunoprecipitation protocol using a
combination of two commercial LRRK2 antibodies to detect endogenous LRRK2 [202,
215, 216]. This method has not been applied to human brain samples or for
quantitative comparison of LRRK2 protein expression.
57
1.3.7 LRRK2 mRNA and protein distribution
In terms of mRNA expression, LRRK2 has consistently been shown to occur
at high levels in kidneys, lungs and lymph nodes [240-243] with lower mRNA
abundance in the brain [244]. In rodent and primate regional brain distribution
studies, the striatum expressed significant levels of LRRK2 mRNA while nigral LRRK2
mRNA appeared to be of considerably lower abundance [245-248]. In an attempt to
correlate LRRK2 protein distribution to early reports of LRRK2 mRNA, relative LRRK2
expression has been described for rodent and primate tissues [244, 248, 249]. LRRK2
antibodies
have
been utilized
for
immunological
detection.
In
general
immunohistochemical data on LRRK2 protein distribution appears to correlate with
the mRNA literature with high LRRK2 abundance in striatum, cortex and cerebellum
[244, 248, 249]. More recent reports, utilizing validated antibodies of better quality
have confirmed high striatal LRRK2 expression providing confidence for localization
and functional data published in earlier studies [250]. In general, LRRK2 brain
distribution involves regions of the dopaminergic pathway. Regional and cell specific
expression may have functional implications for the LRRK2 protein in neuronal subpopulations.
1.3.8 LRRK2 cellular distribution
In addition to tissue distribution studies LRRK2 antibodies have been utilized
to determine LRRK2 cellular localization. Cellular fractionation analyses have
identified LRRK2 in mitochondrial, lysosomal, ER and vesicular compartments further
supported by immunohistochemical and EM studies [201, 218, 251]. LRRK2
dimerization has been confirmed for protein immunoprecipitated from cells with
membrane associated dimers described to exhibit enhanced kinase activity, adding
to the initial in vitro studies [218, 239, 252]. Assuming accuracy in quantitative
immunological assessment there is no evidence to suggest PD associated mutations
58
altered LRRK2 protein levels or cellular distribution in any of the cell models
analyzed [218, 239, 252].
1.3.9 LRRK2 function
LRRK2 detection and characterization of its intracellular distribution in
combination with the phenotypic changes associated with wild type, mutant and
LRRK2 knockdown models have contributed to the functional studies and
investigation of intracellular pathways potentially involving the LRRK2 protein.
1.3.9.1 LRRK2 and vesicles
LRRK2 abundance in synaptosomes and vesicle enriched fractions combined
with changes in dopamine transmission reported for mutant and LRRK2 deficient
mouse models suggested a role for LRRK2 in synaptic function. Knockdown of LRRK2
correlated with morphological abnormalities at the pre synapse [253] and altered
vesicle recycling dynamics [254]. The G2019S hyperactive kinase LRRK2 impaired
synaptic vesicle endocytosis potentially through its association with the vesicular
machinery [255]. The sub-cellular distribution of α-synuclein closely mimics that of
LRRK2 [24]. In addition, α-synuclein has been shown to interact with the soluble Nethylmaleimide sensitive factor (NSF) attachment protein receptor (SNARE)
complex, mediating neurotransmitter uptake and release [256] potentially
suggesting a functional link between α-synuclein and LRRK2.
Another subset of vesicles thought to functionally require LRRK2 are
involved in endocytic-lysosomal trafficking. A membrane association with
multivesicular bodies (MVBs) and autophagic vacuoles (AVs), prime vesicle
components of this pathway, has been demonstrated by electron microscopy [257].
Knockdown of LRRK2 increased flux through the pathway while the R1441C mutant
LRRK2 was described to cause an autophagic imbalance resulting in accumulation of
59
AVs and MVBs. Autophagic clearance of mutant and excess α-synuclein may
therefore involve LRRK2.
1.3.9.2 LRRK2 and microtubules
Axonal degeneration, neurite defects and tau pathology commonly
associated with mutant LRRK2 models and PD patients implicated microtubule
dynamics in PD pathogenesis [258-260]. The subsequent identification of the
interaction between LRRK2 and β-tubulin, the main component of microtubules, and
kinase dependent regulation of tubulin polymerization events [261, 262] supported
a direct mechanistic role for LRRK2 in the observed phenotype. More recently,
another microtubule related protein Tau, the main component of neurofibrillary
tangles in PD and AD, was reported as an integral part of the LRRK2 β-tubulin
complex [263] potentially providing an additional link between LRRK2 mutations and
Tau pathology.
Microtubule regulation has also linked LRRK2 to the canonical Wnt signalling
pathway where a genetic interaction with α-synuclein is thought to take place. Tau
phosphorylation is believed to be enhanced by the interaction of LRRK2 with the
dishevelled protein (Dvl) [264], a downstream target of GSK3β which also induces αsynuclein dependent tau phosphorylation [265].
1.3.9.3 LRRK2 transcription and translation
Multiple protein interaction domains present in the LRRK2 sequence as well
as its cytosolic abundance implicate the protein’s involvement in signalling cascades
transcriptional and translational regulation. Microarray analyses of wild type,
mutant LRRK2 over-expression and knockdown LRRK2 models implicated
transcriptional contribution to the regulation of pathways involved in axonal
guidance, actin cytoskeleton dynamics, calcium signalling, cell cycle and
differentiation as LRRK2 mediated events [266]. Furthermore, mRNA expression of
60
genes involved in neuronal maintenance appeared to be mediated by LRRK2 levels in
mouse embryonic stem cells [267]. The link between dysregulation of gene
expression and PD pathology is demonstrated by increased α-synuclein mRNA levels
in PD patient SNpc [268]. A number of reports suggest LRRK2 regulates α-synuclein
steady state mRNA levels, potentially altering intracellular protein concentrations
which may lead to aggregation [269, 270].
In support of the microarray data, mechanistic analyses implicated LRRK2 in
transcriptional activation, albeit in peripheral tissues whereby LRRK2 was found to
mediate the nuclear factor of activated T-cells (NFAT) cytosolic to nuclear
translocation and gene regulation in response to inflammation [271]. LRRK2
interaction with the dishevelled protein (Dvl) is associated with β-catenin mediated
transcriptional regulation in the canonical Wnt signalling pathway [272].
Translational repression and or degradation of mRNA inhibiting, miRNA
particles through LRRK2 interaction with human Argonaute 2 protein (Ago2) and the
eukaryotic translation initiation factor 4E-binding protein 1 (4E-BP) was shown to
impact on PD associated genes in a LRRK2 kinase dependent mechanism [273]. In
addition aberrant phosphorylation of 4E-BP by mutant LRRK2 correlated with
dopaminergic cell death and morphological defects in the post synapse [231]. LRRK2
was also shown to mediate steady state 4E-BP protein levels suggesting a possible
negative feedback loop [274]. A genetic interaction has been described between
LRRK2, PINK1, parkin and 4E-BP. Knockdown of PINK1 or parkin affected cell viability
of dopaminergic neurons in Drosophila, a phenotype exacerbated by the loss of the
4E-BP protein [275]. Knockdown of LRRK2 rescued this dopaminergic phenotype
suggesting a potential interplay between PD assicated proteins and the 4E-BP
pathway.
61
In addition, LRRK2 has been shown to mediate cellular stress responses
through participation in canonical NFκB and MAPK cascades [203, 276, 277],
pathways previously implicated in PD pathogenesis [278, 279], further emphasising
altered transcriptional response in LRRK2 linked PD.
1.3.9.4 LRRK2 and mitochondria
Mitochondrial involvement in LRRK2 pathogenesis is supported by a number
of pathological, morphological and functional observations. In addition, cellular
localization studies have confirmed a mitochondrial interaction [251, 280, 281]
estimating approximately 10 % of the LRRK2 cellular pool at the outer mitochondrial
membrane. LRRK2 appears to regulate survival of dopaminergic neurons by
conferring resistance to oxidative stress [282] while LRRK2 toxicity induced by
protein over-expression is exacerbated in rotenone and paraquat treated cells [283]
potentially due to increased sensitivity to Complex I inhibition [284]. Mitochondrial
respiratory chain defects have been described in mutant LRRK2 expressing cells
correlating with structural alterations to the mitochondrial network [280, 281, 285].
Dopaminergic cell death has been linked to LRRK2 interaction with the
mitochondrial apoptotic machinery [227, 286].
Although mechanistic links have been proposed for some of the LRRK2
associated cellular pathways with respect to other PD related genes, the strongest
genetic association has involved the mitochondria. Defects in mitochondrial cristae
observed in PINK1 deficient C.elegans were rescued by dLRRK2 knockdown
suggesting an antagonistic mechanism of gene regulation [229]. In addition,
rotenone sensitivity of LRRK2 mutant flies is suppressed by parkin, PINK1 or DJ1 coexpression [284, 287]. Furthermore, excessive mitochondrial fragmentation,
reported in LRRK2, parkin, PINK1 or DJ1 deficient cells can be rescued by the
artificial introduction of mitochondrial fusion machinery Mfn2 or a dominant
62
negative form of DLP1 which promotes fusion [280, 281, 284, 287], potentially
linking the converging mechanisms.
63
1.4 Hypothesis
We hypothesised a role for LRRK2 in the regulation of mitochondrial
function, proposing the G2019S mutation would impact on the electron transport
chain. We predicted, if confirmed, a detailed evaluation of upstream regulatory
pathways could reveal overlapping and novel signalling events potentially reflecting
previously unexplored mechanisms in idiopathic PD.
1.5 Aims
1. To characterize primary and transformed cell culture models expressing
endogenous and ectopic wild type or G2019S LRRK2 protein. Characterization
will require screening of commercially available antibodies for their ability to
detect native, denatured and immunoprecipitated LRRK2 by Western blot and
immunohistochemistry.
2. To determine whether the previously validated immunoprecipitation protocol
can be used for quantitative evaluation of LRRK2 protein levels in order to assess
relative expression levels in the cell models and brain tissue.
3. To use the primary and transformed cells as a LRRK2 model system to look at the
impact of the G2019S mutation on mitochondrial function. A LRRK2 siRNA
knockdown model will be used to address whether the ETC is regulated by
endogenous wild type LRRK2.
4. To determine whether the functional state of the mitochondria in control, wild
type and G2019S LRRK2 over-expressing cells or LRRK2 siRNA knockdown system
correlates with morphological changes in the mitochondrial network.
5. To characterize the effects of LRRK2 kinase inhibitors in the over-expressing
neuroblastoma model and determine the optimal parameters for subsequent
evaluation of LRRK2 kinase contribution to mitochondrial function.
64
6. To explore the functional role of LRRK2 in upstream mechanisms regulating
mitochondrial function.
65
Chapter 2: Materials and Methods
All chemicals were obtained from Sigma or Merck & Co. unless otherwise stated.
2.1 Plasmid constructs
pET-DEST51 plasmids (Invitrogen) containing the cDNA encoding full length
human wild type or G2019S LRRK2 with a C terminal V5 tag, AmpR gene (ampicillin
resistance for bacterial selection) and bsr gene (blasticidin resistance for mammalian
selection) were kindly provided by Mark Cookson (Bethesda, USA), [221].
Mammalian gene expression in pET-DEST51 was driven by the elongation factor 1
(EF-1α) promoter. pEGFPN1 plasmid (Clontech) with an N terminally tagged GFP
UCP2 insert, KmR gene (kanamycin resistance for bacterial selection) was provided
by Koji Nishio (Nagoya, Japan) [288]. Mammalian gene expression in pEGFPN1 was
driven by the cytomegalovirus (CMV) promoter. pGIPZ (Thermo Scientific) vector
(scrambled shRNA: ATCTCGCTTGGGCGAGAGTAAG) or pGIPZ containing shRNA
targeting the 3‘ UTR of the DJ1 gene (CCTACAAATTGTGTCTATA), sequence encoding
GFP, AmpR and Pac (puromycin resistance for mammalian selection) genes were
obtained from the University College London (UCL) shRNA library (Open Biosystems)
and provided by Gyorgy Szabadkai (UCL). Mammalian gene expression in pGIPZ was
driven by the CMV promoter.
2.2 Brain samples
Human brain samples from two control subjects of 24 hour post mortem
delay were obtained from the Queen Square Brain Bank. Cases were collected using
a protocol approved by the London Multicentre Research Ethics Committee and
stored under a license issued by the Human Tissue Authority. Human brain samples
of 5 individuals (3 control, 1 IPD case, 1 Alzheimer’s disease case) of short post
mortem delay times were obtained from Jose Obeso (Pamplona, Spain) and studied
66
with the approval of the Ethics Committee. Details of brain samples are described in
(Table 2.1).
Brain stem, olfactory lobe, mid brain, cerebellum, striatum and cortical
regions were dissected from fresh brains of control male 6 week old mice (C57BL/6
strain). Equivalent tissue regions of one marmoset brain were obtained from Peter
Jenner (King’s College, London). The marmoset brain was from a 2 year old female
animal that received daily subcutaneous MPTP (2.0 mg/kg) injections for 5
consecutive days. LRRK2 knockout mouse brains obtained from C57BL/6 strain
animals with a partial deletion of exon 39 and complete deletion of exon 40 of the
LRRK2 gene [289] were kindly provided by D. Alessi (Dundee).
2.3 Cell lines
Punch skin biopsies from PD patients carrying G2019S LRRK2 mutations
and age matched controls (Table 2.2) were collected by Dr. Daniel G Healy (Royal
Free Hospital, UCL). Human skin fibroblasts were established by Dr. Susannah Horan
and Dr. Jan Willem Taanman (Clinical Neuroscience, UCL).
67
Venous blood (8 mls) was collected from the same subjects with a hypermic needle
as part of a vacutainer in ethylenediaminetetraacetic acid (EDTA) vacuum tubes
(Beckton Dickinson and Co). Peripheral blood mononuclear cells were isolated and
commercially transformed into lymphoblastoid cell lines by Epstein-BarrVirus (EBV)
(ECACC, Salisbury, UK).
SHSY5Y cell lines stably expressing pGIPZ vectors containing shRNA targeting the
3’UTR of the DJ1 gene or scrambled shRNA were generated and kindly provided by
Dr. Zhi Yao (Physiology Department, UCL).
68
2.4 Cell culture
All cell culture medium was supplemented with 10 % (v/v) foetal calf serum
(FCS, BioSera), penicillin (100 U/mL), streptomycin (100 mg/mL) and sodium
pyruvate (100 mg/mL) unless stated otherwise. Adherent cells were grown on 10 cm
plates while suspension cultures were maintained in surface-ventilated 75 cm flasks.
Cells were cultured in a temperature-controlled, humidified incubator with 5 % CO2
at 37 oC (Hera Cell 240, Thermo Scientific, Essex UK).
Adherent cell lines were passaged or harvested by trypsinization in versene
(Invitrogen) containing 0.1 % trypsin enzyme. Briefly, culture medium was aspirated,
cells washed once in phosphate buffer saline (PBS) and 1 ml of trypsin solution
applied to each 10 cm plate with cells monitored until all of the culture detached
from the surface in the trypsin suspension. Trypsin was neutralised with 10 ml
culture medium and cells re-plated at 1:4 dilutions. Unless otherwise stated cell lines
were cultured up to passage 25.
Suspension cells or trypsinized adherent lines were harvested by
centrifugation (1000 g, 10 min, RT). Cell counts were performed using a phase
contrast Olympus CK2 Microscope (London, UK) on Trypan blue stained cells
mounted on Fastread slides (Immune Systems Ltd., Devon, UK).
Cell stocks were frozen slowly to -80 oC in culture medium containing 10 %
dimethyl sulfoxide (DMSO) and stored in liquid nitrogen.
Fibroblasts were cultivated under standard conditions in Dulbecco’s
Modified Eagle Medium (DMEM, Invitrogen) [290]. Skin biopsies were stored in
sterile flasks containing fibroblast culture medium and processed on the same day.
Following removal of subcutaneous fat tissue, skin sections were dissected into
~1mm2 explants and placed skin side up on 5 cm petri dishes containing 2 ml pre
warmed growth medium. Surface tension of the medium maintained biopsies
69
adhered to the plate. Explants were cultured under standard growth conditions with
the addition of 0.5 ml growth medium every 2 days. Upon appearance of epithelial
monolayers and migrating fibroblasts the medium was replaced by 3 mls of fresh
solution. Once confluent, fibroblasts were separated from the epithelial cells by
trypsinization, re-plated in 10 cm dishes and expanded for stocks. Experiments were
restricted to passages 4-12 to avoid cell senescence. In studies performed under
substrate limiting conditions glucose free DMEM (Invitrogen) was supplemented
with 4.5 g/L galactose and cells equilibrated in galactose medium for 24 hours prior
to analysis.
Lymphoblast cells were commercially provided as frozen stocks of 1 million
cells per vial. Cells were defrosted at 37 oC, washed once (1000 g, 10 min, RT) in 10
ml of Roswell Park Memorial Institute (RPMI) 1640 culture medium (Invitrogen)
containing 20 mM HEPES buffer (Invitrogen) and pelleted cells resuspended in 10
mls of fresh growth medium before transferring cell suspensions to 75 cm flasks
[291].
SHSY5Y neuroblastoma cells were cultivated in 50/50 DMEM/F12 medium
(Invitrogen) containing 0.1 % uridine and 1 % MEM non essential amino acids
(Invitrogen) [292].
2.5 Transformation of DH5α cells and DNA extraction
pET-DEST51 and pGIPZ plasmid DNA (approximately 2 µg) was supplied
spotted on Whatman filter paper. Marked regions containing the dried plasmid were
excised under sterile conditions, placed in a sterile eppendorf tube and plasmids
recovered by reconstitution in 50 µl TE buffer (10 mM Tris base, 1 mM EDTA, pH
8.0). Reconstituted plasmids (100 ng) were transformed into competent DH5α
Escherichia coli cells (50 µL, Invitrogen) by incubation on ice (30 min) and heat shock
(2 min, 42 oC) to close the cellular pores. Transformed cells were equilibrated in 1 ml
70
of Lysogeny broth (Sigma) (1 % tryptone, 0.5 % yeast extract, 1 % NaCl, 0.1 %
glucose) on a temperature controlled shaker (37
o
C, 2 hrs, 180 rpm). Isolated
colonies were expanded overnight (37 o C) by plating on 10 cm agar (Sigma) plates
containing appropriate antibiotic (ampicillin or kanamycin at 150 mg/ml where
stated).
Mini plasmid preparations were carried out for DNA sequencing. Single
colonies were inoculated in 5 ml of Lysogeny broth containing antibiotic and
expanded overnight in aerated flasks (37
o
C, 180 rpm). Glycerol stocks (15 %
glycerol, 85 % culture) were prepared for each colony, immediately flash frozen in
liquid nitrogen and stored at -80
o
C for subsequent culture expansions. Plasmids
were isolated from the remaining 4 ml of pelleted culture (4 o C, 10 min, 1000g) with
the Plasmid Miniprep kit (Quiagen, Chatsworth, CA, USA) according to
manufacturer’s instructions and sequenced to confirm the integrated plasmid
product.
Large scale plasmid production (Maxi prep) was required to generate
concentrated plasmid stocks for mammalian cell transfections. Pre cultures in the
exponential growth phase (10 hrs, 37
o
C, 180 rpm) were transferred to aerated
conical flasks containing 50 ml broth with appropriate antibiotic and expanded for a
further 8-10 hours. Plasmids were isolated from pelleted cultures (4
o
C, 10 min,
1000g) with the Wizard miniprep kit (Promega, Hampshire, UK) or Maxiprep kit
(Quiagen, Chatsworth, CA, USA) according to manufacturer’s instructions. The
Wizard miniprep kit was specifically used for purification of LRRK2 containing pETDEST51 plasmids as it generated the best yield for this plasmid (1 mg of plasmid per
10 ml of culture). Quiagen maxiprep kit was used to purify the pEGFPN1 plasmid
containing the UCP-GFP construct.
71
2.6 Concentration and purity of nucleic acids
Concentration and purity of DNA and RNA samples was determined using a
NanoDrop 1000 spectrophotometer (Thermo Scientific). Briefly, 1.5 µL of the sample
was spotted onto the pedestal of the machine with sample absorbance determined
at 260 nm (nucleic acid) and 280 nm (protein). The concentration of nucleic acids
was calculated electronically by the NanoDrop 1000 software based on the equation
A=εcl; where A is the absorbance of the sample, ε is the extinction coefficient, c is
the concentration and l is the path length of sample. The 260/280 absorbance ratio
was subsequently used to determine the purity of the DNA/RNA sample whereby
values >1.8 for DNA and >2.0 for RNA were acceptable. Lower ratios suggested the
presence of protein, phenol and other contaminants.
2.7 Restriction enzyme digest of pET-DEST51
To increase the probability of integration of the full length LRRK2 gene in
mammalian cells LRRK2 containing plasmids were enzymatically digested targeting a
region outside the LRRK2 coding sequence. The Fsp1 restriction enzyme (restriction
site 5’tgc^gca3’, 3’acg^cgt5’, 5000 U/ml, New England Biolabs, Hitchin, UK) was
chosen targeting position 6886 of the 7464 base pair pET-DEST51 plasmid. pETDEST51 plasmids (25 µg) purified as described in section 2.5 were digested (37 oC, 2
hours) with Fsp1 in a final volume of 200 µl containing 20µl NEB buffer 2 (New
England Biolabs). Digestion was terminated by enzymatic inactivation (65 oC, 20
min). Digested constructs were recovered by ethanol precipitation (10 % NaAc, 250
% v/v ethanol, -20 oC, o/n) and pelleted by centrifugation (4 oC, 30 min, 17,000 g).
Pellets were washed in 80 % ethanol, air dried and resuspended in 20 µl deionised,
sterile TE buffer.
72
2.8 Agarose gel electrophoresis
Electrophoresis of PCR products, intact and digested plasmids was carried
out on agarose gels. Briefly, 1 µg of sample was combined with 6X loading buffer
(Fermentas) and electrophoresed (60 V) on 0.8 – 2 % Agarose gels prepared in Trisacetate-EDTA buffer ((TAE), 40 mM Tris, 20 mM acetic acid, 1 mM EDTA, pH 8)
containing 0.005 % SYBR safe DNA gel stain (Life technologies) for visualization of
separated band products. Bands were visualised by ultraviolet illumination with
relative sizes estimated based on migration of adjacent reference markers (BioLabs).
UV illuminated bands were captured on an Olympus digital camera.
2.9 Transfection of mammalian cells
SHSY5Y cells stably expressing wild type and G2019S LRRK2-V5 containing
pET-DEST51 plasmids were generated and kindly donated by Dr. JM Cooper and Dr.
K Chau (Clinical Neurosciences, UCL). Transfections of SHSY5Y cells with the
linearized pEF-DEST51 plasmids containing the LRRK2 gene were carried out as
previously described [293] using Superfect reagents (Quiagen, Hilden, Germany)
according to manufacturer’s instructions. Plasmids (25 µg) were complexed with
transfection reagent by incubation (15 min, RT) in serum free cell culture medium.
Serum was omitted to prevent potential interference with complex formation.
SHSY5Y Cells grown at 40 % confluency on 10 cm plates were treated with the
mixture for 4 hours allowing for endocytosis of the plasmid, cultured overnight
under standard conditions and subjected to blasticisdin (400 µg/ml) or puromycin
selection (400 µg/ml) over a 6 week period. Culture medium was replaced every 2
days to remove dead cells with the formation of antibiotic resistant colonies
monitored over this time course. Isolated colonies grown to a radius of 0.5 mm were
separated by polystyrene microchambers (Sigma), trypsinized and transferred to 2
cm plates. Clones were grown to confluency and expanded for frozen stocks.
73
For transient protein expression purified plasmids were introduced into
cells by electroporation. Cells (2.5 x 105) were harvested, washed once in PBS and
resuspended in a mixture of 9 µl of reagent R and 1 µl of plasmid DNA (1 µg/ µl). The
mixture was electroporated (Neon, Invitrogen, 1000 V, 30 seconds, 2 pulses),
transferred to 1 ml growth medium without antibiotic and plated (1:5) on 6 well
plates in a final volume of 2 ml. Cells cultured for a further two days under standard
growth conditions.
Gene silencing in SHSY5Y cells was achieved by introduction of small
interfering RNA (siRNA) by the method of transient transfection. Details of siRNA
sequences are described in (Appendix Table A1). Optimal siRNA delivery conditions
for SHSY5Y cells were determined using fluorescently labelled siRNA constructs
(Ambion). Transfection reagents; HiPerfect (Qiagen, Hilden, Germany), TransIT-TKO
(Mirus ML015), TransIT-siQUEST (Mirus ML033), TransIT-2020 (Mirus MIR5400) were
tested to determine which reagent achieved the most efficient siRNA delivery in the
SHSY5Y cell line. Two volumes of transfection reagent (1 and 4 µl) were tested to
determine whether the concentration would affect the efficiency of siRNA delivery.
Transfection reagent was incubated with 10 nM fluorescently labelled siRNA in a
final volume of 100 µl serum free culture medium (15 min, RT). SHSY5Y cells were
plated at a density of 9 x 10-4 per well (24 well plate) in 500 µL of complete culture
medium. Transfection mix was immediately added to the plated cells which were
subsequently cultured overnight under standard culturing conditions. Transfected
cells were identified by immunofluorescence 24 hours post transfection by red
fluorescent punctate staining indicating the uptake of siRNA (Fig 2.1).
Red punctuate staining was observed for SHSY5Y cells transfected with 4 µL
of HiPerfect, ML015 and MIR5400 transfection reagents (Fig 2.1). 4 µL of HiPerfect
reagent were chosen as the optimal siRNA delivery conditions as all the transfected
74
SHSY5Y cells contained red punctuate structures, unlike ML015 and MIR5400 where
30 % and 60 % of the cells were positive for red fluorescence.
2.10 Reverse transcriptase
Total RNA was extracted from fibroblasts or SHSY5Y cells (1-5x104) with the
RNeasy Mini kit (Qiagen) according to manufacturer’s instructions and quantified
spectraphotometrically as described in Section 2.6. Purified RNA template (0.5 µg)
was denatured and annealed (65 oC, 5 min) to random nanoamine primers (1 µL) in
final volume of 10 µl (Primer design). Samples were cooled on ice (5 min) prior to
addition of deoxyribonuclease 1 (1 µL), deoxynucleaotide triphosphate mix (dNTPs,
0.5 mM), reducing agent dithiothreitol (DTT, 10 mM) (Primer Design) and sterile
water to a final volume of 20 µl. The reverse transcriptase extension step (25 oC, 5
min, 55 oC, 20 min), initiated at the primer binding sites, reverse transcribed the RNA
template to produce complementary DNA (cDNA). Following reverse transcription,
the enzyme was heat inactivated (75 oC , 15 min).
75
2.11 Real Time PCR
For quantification of gene expression, 1 µL of the first strand cDNA sample
was used as template in quantitative real-time PCR amplification with the TaqMan
system or SybrGreen assay (Applied Biosystems, Paisley, UK). In TaqMan reactions
an initial polymerase activation step (95 oC, 10 min) was followed by 40 cycles of PCR
as described in the supplier’s protocol (15 sec at 95 oC, 1 min at 60 oC). Annealing
temperatures for SybrGreen primers are described in (Appendix, Table A2),
determined by subtracting 5 oC from the melting temperature (Tm) of the primer
strand as calculated by the manufacturer (Eurogenetech), based on the length and
nucleotide composition of the primer. Standard denaturation (95 °C, 15 sec) and
extention (72
o
C, 30sec) conditions were maintained across all SybrGreen
experiments.
SybrGreen primers were validated by cDNA template titration with a valid primer
concentration range for PCR requiring one amplification cycle increase with doubling
of template concentration (Fig 2.2A). RT PCR product melting curves were generated
in a temperature range of 60-90 oC at 0.3 oC increments for each SybrGreen reaction
to ensure specificity of amplifications for the gene target. Melting points of
amplicons were considered appropriate in the range from 80 oC to 85 oC with
additional peaks in the lower temperature range (60-80 oC) assigned to primer
dimers (Fig 2.2B, C). Furthermore, TaqMan and SybrGreen PCR product sizes were
checked by agarose gel electrophoresis with relative sizes estimated based on
migration of adjacent reference markers (BioLabs) (Appendix, FigA2).
The comparative cycle threshold (CT) method was used to quantify mRNA
expression, correcting gene amplification to GAPDH CTs determined by the
equation: Relative mRNA expression = (2-[∆∆Ct], ∆∆Ct = Ctgene-CtGAPDH) where the Ct
value is the cycle number at which the fluorescence signal crosses the threshold
76
(0.01), (Fig 2.2D). Threshold values were chosen based on the first detectable
increase in fluorescence from RT-PCR products and maintained across all
experiments.
77
2.12 LRRK2 mutation screen in fibroblasts and SHSY5Y cells
cDNA from control or G2019S fibroblasts, wild type and G2019S LRRK2-V5
over-expressing SHSY5Y cells was screened for the presence of the G2019S mutation
at position 6055 of the LRRK2 gene. Primers were designed targetting the LRRK2
sequence flanking a 700 bp fragment on exon 41 of the LRRK2 gene containing the
6055 nucleotide as part of the codon encoding the G2019A/S residue (Table A2).
Primers (0.5 µM) were combined with 2 µL of total cDNA, and Phusion Flash High
Fidelity PCR mastermix (Thermo Fisher Scientific, Loughborough, UK) in a final
volume of 50 µl. The targeted region of LRRK2 was amplified by 30 cycles of a three
step PCR amplification protocol as suggested by the manufacturer (denaturation 98
o
C 10 min, annealing 20 sec, elongation 72 oC 30 sec). Annealing temperatures were
determined by subtracting 5 oC from the melting temperature (Tm) of the primer
strand as calculated by the manufacturer (Eurogenetech), based on the length and
nucleotide composition of the primer. A final 10 minute extention step (72 oC)
completed the PCR reaction. The approximate size of PCR products was determined
by agarose gel electrophoresis (Section 2.8). Products were sequenced at the
Scientific Support Unit (Wolfson Institute for Biomedical Research, University
College London). Sequence comparison was ncarried out by BLAST alignment
(PubMed) on the human LRRK2 fasta entry registered in the Uniprot Database
(NP_940980) with sequence profiles generated by Sequencher 5.0.
2.13 Whole cell and tissue extractions
SHSY5Y, fibroblast, lymphoblast cell pellets (1x106) or brain tissue (1 mg
dry weight) were prepared for Western blot analysis by solubilization in 50 µl 0.1 %
sodium dodecyl sulphate (SDS), 10 mM Tris HCL containing a cocktail of protease
inhibitors (Thermo Scientific, targeting serine, cysteine, aspartic acid-proteases and
amino peptidases). Brain tissue was homogenized in buffer with a Potter type
78
homogeniser, cleared by centrifugation (4oC, 17,000 g, 30 min) and sonicated (2 x 15
s pulses on ice (Decon Fs 100b)). Chromosomal DNA in whole cell extracts was
digested by treating samples with DNAse1 (Promega, 1/10th the total reaction
volume) in enzyme buffer (1/10th the total reaction volume, 45 min, 37 oC).
Following DNAse treatment cell extracts were centrifuged (4oC, 17,000 g, 10 min) to
remove insoluble and aggregated components.
2.14 Immunoprecipitation
Immunoprecipitation procedures (IP) essentially followed those published
by [215]. Cell pellets (6x106 cells per 500 µl buffer) were extracted (10 min, on ice) in
lysis buffer (LB: 50 mM Tris HCl, 0.27 M sucrose, 1 mM Na3VO4, 1 mM EDTA, 1 mM
EGTA, 10 mM β-glycerophosphate, 5 mM sodium pyrophosphate, 50 mM NaF
supplemented with 1 % Triton X100 (IP) or 1 % nonidet P-40 (NP-40) (for co-IP), 1
mM benzamidine, 0.1 mM phenylmethanesulfonylfluoride (PMSF) and protease
inhibitor cocktail), retaining detergent soluble fractions following centrifugation (4
o
C, 17,000g, 10 min).
For brain IP, samples were processed as described in the Abcam IP
datasheet in LB buffer (2.5 mg frozen tissue per 1 ml buffer) by 30 strokes with a
Potter type homogenizer, pulsed twice by sonication and cleared by centrifugation
(4oC, 17000 g, 30 min). In co-IP experiments sonication steps were substituted by 10
passages through a 27 gauge needle (Terumo) to maintain integrity of protein
complexes.
Immunoprecipatation required 1 mg of protein, incubated with 20 µL of
Protein G agarose beads (Invitrogen) pre coupled to antibody (4 °C, 30 min). 1 mg of
lysate with 20 µ agarose beads without the addition of antibody were included in
parallel control experiments to discount potential binding of immunoprecipitated
proteins to beads. Samples were incubated with rotation (4 °C, 2 hours), beads
79
collected by centrifugation (17,000 g, 1 min) and non specific interactions removed
by washes in salt (500 mM NaCl for IP, 150 mM for co-IP) containing LB followed by
two washes in LB buffer (17,000 g, 1 min). Pulled down proteins were eluted from
the beads by solubilisation in 50 µl 2 X LDS sample buffer (Invitrogen) supplemented
with 2X reducing agent (Invitrogen) with 25 µl of the eluted sample used for SDS
PAGE.
An un-cropped Western blot of LRRK2 immunoprecipitates from control
fibroblasts is shown in (Appendix Fig A3) where a single 280 kDa 3514-1
immunoreactive species was observed.
2.15 Protein quantification
Protein quantification was carried out using the bicinchoninic acid (BCA)
protocol (Pierce, Cramlington, UK) [294]. 2 and 4 µl of SDS (Section 2.13) or Triton
X100 (Section 2.14) cell and tissue extracts were mixed with PBS to a final volume of
25 µl in a 96 well plate (Starsted). Blank samples were those containing sample
extraction buffer and PBS while bovine serum albumin (BSA, 0-2000 µg/ml) was used
as protein standard. Kit components A and B were mixed in a 1:50 ratio and 200 µL
of the mixture added to each well. Samples were incubated at 37 oC for 30 minutes
and assessed spectraphotometrically by quantifying the light absorbance at 562 nm
(Synergy HT BioTEK spectrophotometer). Protein concentrations were determined
from the linear absorbance plot generated by the known concentrations of bovine
serum albumin (BSA) protein standards for each measurement.
2.16 SDS PAGE
Western blot analysis of SDS soluble lysates (Section 2.13) or Triton X100
extracts (Section 2.14) required 25 µg of protein denatured in NuPAGE LDS sample
buffer (Invitrogen) supplemented with NuPAGE sample reducing agent (Invitrogen).
Proteins were heated (65 °C, 5 min) and electrophoresed on precast NuPAGE Novex
80
12 % or 4-12 % polyacrylamide Bis-Tris gels (Invitrogen) in the presence of MES (50
mM MES, 50 mM Tris Base, 0.1% SDS, 1 mM EDTA, pH 7.3 for higher molecular
weight proteins, >150 kDa) or MOPS (50 mM MOPS, 50 mM Tris Base, 0.1% SDS, 1
mM EDTA, pH 7.3, for lower molecular weight species, <150 kDa) SDS running buffer
(Invitrogen) (RT, 45 min, 200 V) using the BioRad system (Hertfordshire, UK). For
proteins of molecular weight >150 kDa, electrophoresis was extended to 60 minutes
to improve the separation of higher molecular weight species [244].
2.17 Western blot analysis
SDS PAGE separated proteins were transferred to polyvinylidene fluoride
(PVDF) membranes (GE Healthcare) (RT, 90 min, 30 V, BioRad system). Transfers
were extended to 120 minutes for higher molecular weight proteins (>150 kDa).
Membranes were subsequently blocked (RT, 1 hr) with 5 % w/v semi skimmed milk
containing PBS with tween (PBS-T 0.3 % v/v) and incubated with primary antibodies
under specified conditions (Appendix Table A3). Following washes in PBS-T (3 X 10
min), horseradish peroxidase labelled secondary antibodies (Dako) were applied at
1:3000 dilution in 5 % milk (RT, 1 hr). Membranes were washed again with PBS-T and
protein bands visualized by chemiluminescence (Thermo Pierce) on X-ray film (GE
Healthcare) estimating band molecular weights relative to standard protein markers
of 10-250 kDa range (Precision Plus Protein Standards, BioRad). Protein bands were
quantified by densitometry using Image J software.
2.18 Native gel electrophoresis
Blue native PAGE (BN PAGE) electrophoresis was carried out to investigate
the native structure of LRRK2 complexes, according to manufacturer’s instructions
based on the protocol developed by [295]. To recover soluble proteins, flash frozen
brain tissue (1 mg) or freshly prepared cell pellets (8x106 cells) were homogenised
(Potter homogeniser) in 50 µl volume of 4X Native PAGE sample buffer (50 mM
81
BisTris, 6N HCl, 50 mM NaCl, 10 % w/v glycerol, 0.001 % Ponceau S, pH 7.2,
Invitrogen) with the addition of a cocktail of protease inhibitors (Thermo scientific).
Samples were further processed with a 27 gauge needle (20 strokes) and cleared by
centrifugation (4oC, 30 min, 20,000 g). Protein samples (20 µg) were loaded on to 412 % Native-PAGE Novex Bis-Tris gels (Invitrogen) in the presence of 0.25 % v/v
coomassie blue loading additive (Invitrogen). Electrophoresis was carried out in
anode buffer (AB, outer tank chamber: 50 mM BisTris, 50 mM Tricine pH 6.8) and
cathode buffer (inner tank chamber, AB containing 0.02 % v/v coomassie brilliant
blue) at 150 V. Once approximately 1/3 of the coomassie blue border migrated
down the gel, the cathode buffer was replaced by AB containing 0.002 % v/v
coomassie buffer and electrophoresed for a further 1 hour (150 V). Separated
proteins were transfer to PVDF membranes (30 V, 120 min). Membranes were
equilibrated in absolute methanol to remove excess coomassie dye. NativeMark
Unstained Protein Standards (20-1236 kDa, Invitrogen) were visualized by incubating
in Ponceau S staining solution as reference markers to approximate the molecular
weight of resolved protein complexes.
2.19 Immunofluorescence of cultured cells
Cells plated up to 40 % confluency on 22 mm coverslips were fixed in 4 %
paraformaldehyde (PFA, room temperature, 15 min) and permeabilized in absolute
methanol (MeOH, -20 oC, 20 min) [296]. For visualization of nuclear proteins,
paraformaldehyde fixation was followed by permeabilization with 0.1 % Triton
(room temperature, 30 min) [297]. Nuclear staining required detergent
concentrations to be maintained throughout the immunostaining procedure.
Immunocytochemistry for LRRK2 labelling omitted PFA and required MeOH (-20 oC,
20 min), acetone (-20 oC, 5 min) permeabilization/fixation conditions [298]. During
protocol optimization to determine optimal conditions for LRRK2 staining; PFA
82
fixation followed by sodium citrate (95 oC, 20 min), 1 % Triton X 100 (RT, 30 min) or
MeOH permeabilization (as above) were also tested on control versus wild type
LRRK2-V5 over-expressing SHSY5Y cells. Permibilized cells were subsequently
processed in parallel using the same conditions.
PBS washes (3X times) were preformed after each stage of the staining
procedure. Following permeabilization, non specific binding was blocked by 10 %
normal goat serum (NGS, Life Technologies) (RT, 30 min) and primary antibodies
applied at appropriate dilutions (Appendix, Table A3) in 10 % NGS (RT, 2 hrs),
followed by incubation with fluorescently labelled secondary antibodies (1:200,
Molecular Probes) in 2 % NGS. Coverslips were mounted on glass slides with Citifluor
containing Hoechst trihydrochloride trihydrate (1 mg/ml, Invitrogen) for nuclear
counterstain.
Fluorescent images were acquired using the same exposure settings with an
Axiophot fluorescent microscope, KS400 software (Zeiss, Welwyn Garden City, UK)
equipped with 40X and 100X PlanNeoFluar lenses and the appropriate fluorescent
filters.
2.20 Immunohistochemistry on human and mouse brains
Flash frozen mouse brain hemispheres embedded in optimum cutting
temperature medium (OCT, Sakura) were cut into 12 µm sagittal slices by cryostat
(OTF5000, Bright, Huntingdon, UK) sectioning and mounted on glass slides. Sections
were dried (37 oC, 15 min) and frozen (-80 oC) until further processing. On the day of
the staining, sections were thawed (RT, 15 min) and fixed in 4 % PFA (RT, 15 min)
[299].
PBS washes (3X, 5min) were preformed after each stage of the procedure.
Endogenous peroxidases were blocked (MeOH, 1 % H2O2, 10 min), followed by
blocking of non specific binding (10 % goat serum, 30 min) and incubation with
83
primary antibodies (Table A3, 4 oC, o/n). The following day, sections were incubated
with biotin labelled anti-rabbit secondary antibodies (Gibco, 1:200, RT, 30 min) and
positive immunoreactivity visualized by 3’,3’diaminobenzidine (DAB) oxidation
induced by hydrogen peroxide (1% in PBS solution). DAB brown colour development
was monitored at 1 minute intervals to obtain the optimal signal to noise staining for
each set of experiments. DAB oxidation was terminated by washing slides in PBS
solution. Nuclei were counterstained with haematoxyline (5 min) and differentiated
in acid alcohol (0.3 %) to remove background colouration. Differentiation was
arrested by rinsing slides in warm tap water where the alkaline environment
converted the histone bound haematin to a blue colour. Sections were dehydrated
in ethanol baths of increasing ethanol concentration (70 – 100 %) followed by xyline
to prepare slides for mounting with organic based neutral mounting medium (DPX,
Leica).
Images of relevant brain regions were captured on the Leica opitical imaging
system (DM4000 B, Wetzlar, Germany) at 20X, 100X and 200X magnifications.
2.21 LRRK2 kinase activity assays
In vitro LRRK2 kinase activity assays were performed on LRRK2
immunoprecipitated from 1 mg of cell lysate or 0.5 µg of recombinant GST-LRRK2
(with the help of Francisco Inesta-Vaquera (Dundee)). Recombinant GST-LRRK2, GSTMoesin and GST-UCP2 proteins were prepared and purified by Francisco InestaVaquera (Dundee) as previously described [202]. Briefly, 5 µg of pEBG-2T encoding
residue 1326-end of LRRK2, full length Moesin or full length UCP2 was transiently
transfected into HEK293 cells with 60 µl polyethylenimine in 3 mls of media, as
suggested by the manufacturer (Sigma). Cells were cultured for 36 hours following
transfection and extracted in LB (50 mM Tris HCl, 0.27 M sucrose, 1 mM Na3VO4, 1
mM EDTA, 1 mM EGTA, 10 mM β-glycerophosphate, 5 mM sodium pyrophosphate,
84
50 mM NaF) supplemented with 0.5 % NP40 and 150 mM NaCl. GST tagged proteins
were purified from cell lysates by chromatography on glutathione-sepharose and
bound proteins eluted in buffer containing 20 mM glutathione and 0.27 M sucrose.
Kinase assays [215] were carried out in a volume of 50 µL containing kinase
reaction buffer (RB, 50 mM Tris, 0.1 mM EGTA, pH 7.5), 10mM MgCl2, 0.1 M DTT, 20
µM of LRRKtide substrate (RLGRDKYKTLRQIRQ) and 0.1 mM [32P] adenosine
triphosphate (ATP) (300 cpm/pmol). LRRK2 bound to IP beads was in a volume of 15
µL with the final reaction volume obtained by the addition of deionised H2O.
Reactions were initiated with the addition of the LRRK2 kinase and incubated for 20
minutes at 30 oC. Reactions were terminated by addition of 12.5 µL EDTA (50 mM),
centrifuged (17,000 g, 1 min) and 40 µL of the supernatant spotted on
phosphocellulose p81 paper (Whatman). Samples were air dried and washed
sequentially in 50 mM ortho-phosphoric acid (20 min, 3 times). Following acid
treatment, samples were rinsed with 100 % acetone, dried and the amount of
incorporated radioactivity assessed by Cerenkov counting in the presence of
scintillation fluid.
The LRRK2 kinase specific activity was calculated as: SA= cpm from
reaction/[(SA of [32P] ATP in cpm/pmol)*(reaction time (min))*(enzyme amount in
µg)]*[(reaction volume)*(spot volume)].
LRRK2 bound to beads in the pelleted fraction was eluted by addition of 4X
sample buffer (Invitrogen) and analysed by Western blotting to correct for LRRK2
input.
2.22 Isolation of PMBCs and cell sorting
Peripheral blood mononuclear cells (PMBCs) were isolated from 50 ml of
whole blood by density gradient centrifugation (Axis Shield, Dundee, UK), as
described by the manufacturer. Briefly, blood (diluted 1:1 in isolation buffer (IB: PBS,
85
0.5 % bovine serum albumin (BSA), 2 mM EDTA, pH7.2), layered onto a buffy coat
gradient was fractionated by centrifugation (8,000 g, RT, 20 min). The white PMBC
enriched layer located below the serum fraction was extracted with a 3 ml pasteur
pipette into a 50 ml falcon tube. PMBCs were subsequently washed in cold IB (4 oC,
800 g, 5 min), counted as described in section 2.4 and incubated with magnetically
labelled anti CD4, CD8 or CD19 antibodies (10 µl/105 cells) (Miltenyi, Surrey, UK, 4
o
C, 1hr). Labelled cells were purified by positive magnetic selection (MiniMACS
separator, Miltenyi) on LS columns (Miltenyi).
Purified cells were re-suspended at a concentration of 1x106cells/ml in HBSS
(156 mM NaCl, 3 mM KCl, 2 mM MgSO4, 1.25 mM KH2PO4, 2 mM CaCl2, 10 mM
glucose, 10 mM HEPES, pH 7.35) and 100 µl of the cell suspension added to 4ml
polystyrene round bottomed tubes (Sigma). Mouse anti human conjugated
monoclonal antibodies used to analyse purified cells were supplied by Becton
Dickinson Biosciences (Oxford, UK) and used at a volume of 10 µl/105 cells. 10 µl of
CD4 (FITC), CD8 (FITC), CD19 (PC) and CD21 (CytC) were added to the tubes, samples
vortexed and incubated at room temperature in the dark for 15 minutes. Cells were
then washed free of excess antibody by adding 2mls of HBSS and centrifuged (200 g,
5 min). Supernatants were decanted and cells re-suspended in 500 µl of HBSS prior
to flow cytometry acquisition.
FACS analysis was carried out by Janet North (Royal Free Hospital, UCL). All
readings were acquired on a BD FACSAria II analyser as list mode data files using
FACSDiva software. A forward scatter (FSC) versus side scatter (SSC) dot plot was
created to gate on the live lymphocyte like region of cells (P1) and to exclude cell
debris. A second dot plot was created to show expression of cells from P1 region for
CD19 against FSC. Region P2 was drawn to include all live cells from P1. A third dot
plot was generated to show CD19 versus CD21 which included cells from P1 and P2
86
regions. A fourth dot plot was created to look at the expression of CD4 and CD8 on
P1 and P2 regions to identify any contaminating T cells. Purity of CD8 and CD4 T cells
was analysed in a similar manner. Cell populations of >90 % purity were used for
Western blot analysis.
2.23 Cellular fractionation
Lymphoblasts or SHSY5Y cells were harvested as described in Section 2.4.
Cell pellets (20x106) were resuspended in homogenization buffer (2 ml, HB: 320 mM
sucrose, 10 mM Tris HCl, 1 mM EDTA pH 7) containing protease inhibitors (Thermo
Scientific). Disruption of plasma membranes was achieved by nitrogen cavitation
(Parr Instrument Company, Moline Illinois) at 1200 psi (4 oC, 20 min). Sub-cellular
fractions were obtained by sequential centrifugation at; 1000 g (P1); 12,000 g (P2);
27,000 g (P3) (4 oC, 10 min, Sorvall Super T21, DuPont); 47,000 g (P4) (4 oC, 15 min,
Kontron, P4, T124, Centrikon, Zurich, Switzerland) and 200,000 g (P5) (4 oC, 90 min,
P5, Beckman Coulter, High Wycombe, UK) respectively [300]. Pellets were washed
once in HB buffer while cytosolic fractions were concentrated in 30 kDa cut off
concentrator columns (Millipore) to a final volume of 50 µL. Fractions were
solubilised in 50 µl of SDS sample buffer and 10 % of each sample analysed by
Western blotting.
2.24 Affinity purification of mitochondria
Lymphoblasts or SHSY5Y cells were harvested as described in Section 2.4.
Cells (4x106) were lysed by repeated passage through a 27 gauge needle. Optimal
homogenization conditions were determined by visual inspection of intact cell
densities in trypan blue stained cell sample following increasing strokes with a 27
gauge needle in 1 ml buffer supplied with the mitochondrial isolation kit (Miltenyi,
Surrey, UK). Conditions (25 strokes for lymphoblasts, 30 strokes for SHSY5Y cells)
achieving >90 % disruption of cells were chosen for subsequent experiments.
87
Extracted cell suspensions were centrifuged to remove unbroken cells and nuclei (10
min, 4 oC, 1000 g). 10 % of the post nuclear supernatant (PNS) was retained to
evaluate mitochondrial recovery. Magnetically tagged Tom22 antibodies labelled
mitochondria in cell extracts (4 oC, 1 hr with rotation) and immunoreactive products
were isolated by positive magnetic selection (MiniMACS separator, Miltenyi) on
isolation columns (LS columns with a 30 µM cut-off, Miltenyi) [301]. Magnetically
bound products were washed three times in buffer supplied with the kit and eluted
in a final volume of 1 ml by removing samples from the magnet and applying the
plunger supplied with the column. Affinity purified mitochondria were pelleted by
centrifugation (12,000 g, 2 min).
The PNS sample and affinity purified mitochondria were solubilised in
equivalent volumes of SDS sample buffer and analysed by Western blotting as 1/10
PNS and 9/10 mitochondria.
2.25 Mitochondrial sub-fractionation
Mitochondrial
enriched
fractions
(MEF)
isolated
by
differential
centrifugation (12,000 g, as described in section 2.23) or mitochondria isolated by
affinity purification from (4x106 for each extraction condition) SHSY5Y cells were
incubated (30 min, 4 oC) in HB buffer, HB supplemented with increasing
concentration of digitonin (0.5 – 3 mg/ml), hypotonic buffer (10 mM Tris HCl, 100
mM NaCl, pH 7.6) or 100 mM Na2CO3 (pH 11.5 followed by 5 min sonication) [302].
Samples were separated into soluble and insoluble components by centrifugation (4
°C, 20,000 g, 30 min) and pellets washed in appropriate buffers. Final pellets and
soluble fractions were solubilised in 50 µl of SDS sample buffer and 10 % of each
sample analysed by Western blotting.
88
2.26 Proteinase K digestion of mitochondria
Affinity purified mitochondria (12x106 SHSY5Y cells) resuspended in
proteinase K buffer (0.6 M sorbitol, 20 mM HEPES K-OH, pH 7.4) were digested with
0.5 or 1 µg/ml freshly prepared proteinase K enzyme (stock 50 µg/ml in buffer PB)
for increasing incubation periods (0-30 min) [302]. Digestion was terminated by
inhibiting proteinase K with phenyl methyl sulphonyl fluoride (PMSF, 1 µM, 4 °C, 10
min). Samples were subsequently separated into soluble and insoluble components
by centrifugation (4 °C, 17,000 g, 10min), solubilised in 50 µl of SDS sample buffer
and 10 % of each sample analysed by Western blotting.
2.27 Single cell analysis by confocal microscopy
2.27.1.1 Mitochondrial membrane potential: TMRM
Tetramethylrhodamine Methyl Ester (TMRM) is a cationic lipophilic
fluorescent dye incorporated into the mitochondrial matrix specifically by the
electrochemical gradient generated by the mitochondrial membrane potential (Ψm)
and can be used to quantify the Ψm in live cells. Fibroblasts or SHSY5Y cells were
seeded on 20 mm coverslips at 30 % confluency and cultured over night under
standard culturing conditions. The following day, mitochondrial membrane potential
was assessed by staining cells with 25 nM TMRM (Molecular probes, 45 min, RT) in
respiration buffer (HBSS: 156 mM NaCl, 3 mM KCl, 2 mM MgSO4, 1.25 mM KH2PO4, 2
mM CaCl2, 10 mM glucose, 10 mM HEPES, pH 7.35) [303]. TMRM intensity was
measured in the presence of oligomycin (2 mg/ml), trifluorocarbonyl cyanide
phenylhydrazone (FCCP, 0.5 M), LRRK2 IN-1 inhibitor (1 µM 90 min treatment prior
to rate measurements) and genipin (1.375-10 nM) where stated.
Fluorescent images were acquired on an inverted laser scanning confocal
microscope (LSM510 Zeiss, Oberkochen, Germany) by exciting samples at 543 nm
with maximum laser power directing speciments with the HFT 485/540 short pass
89
filter. Emitted red fluorescence was captured by 40x magnification through an oil
objective with numerical aperture of 1.2 where a 560 nm long pass filter transmitted
all light above 560 nm. Pinhole diameter was set at 244 to eliminate out of focus
fluorescence reaching the detector. The detector gain was adjusted to 464 allowing
for amplification of the fluorescent image and the amplifier offset set to 0.035 as the
cut off thresh hold below which pixels were displayed as black. 12 Bit images were
collected as the average of four scans (scan speed 7); 255 x 255 pixels for normal
resolution increasing the pixel depth to 1024 x 1024 for high resolution images. In
time course measurements images were acquired by a single image scan as 8 bit;
255 x 255 pixels at 5 or 8 minute intervals to reduce laser exposure to cells.
Images were projected through the 1.52 µm Z stacks for a maximum
intensity two dimentional TMRM image. TMRM fluorescence was quantified using
Metamorph software calculating the average signal intensity per cell with a set
threshold value to eliminate background noise determined independently for the
daily set of TMRM measurements. Approximately 30 individual SHSY5Y cells or 15
fibroblasts were analysed for each imaged field, four fields per coverslip in two
(SHSY5Y) or four (fibroblasts) independent experiments (n=200 cells per cell line).
Data was expressed as the mean ± SEM, relative TMRM fluorescence (RFU) per cell.
2.27.1.2 Method validation for TMRM
TMRM staining of control fibroblasts resulted in accumulation of the red dye
within the mitochondrial compartment ((↓)Fig 2.3A). The TMRM fluorescence was
found to be stable for at least one hour after the initial incubation period and all
subsequent measurements were confined to this time frame (Fig 2.3A).
Trifluorocarbonyl cyanide phenylhydrazone (FCCP) is a chemical uncoupler of
oxidative phosphorylation known to dissipate the mitochondrial membrane
potential. FCCP addition (1 µM) resulted in a 2 fold reduction of the TMRM signal
90
intensity accompanied by redistribution of the dye to the cytosolic compartment
(*↓) consistent with dissipation of the proton gradient (Fig 2.3A, B). TMRM
response to FCCP treatment validated the specificity of the dye for mitochondrial
structures in pre treated cells.
Oligomycin inhibits the mitochondrial ATPase, complex V of the
mitochondrial respiratory chain resulting in the accumulation of protons in the
mitochondrial matrix and increased mitochondrial membrane potential. Monitoring
TMRM intensity over time in the presence of oligomycin (4 mg/ml) detected a
gradual 75 % increase in TMRM fluorescence when compared to the stable signal of
untreated cells (Fig 2.3C, D). Increased TMRM intensity reflected changes to the
mitochondrial proton gradient consistent with the accumulation of protons in the
mitochondrial matrix due to inhibited ATPase activity. TMRM response to FCCP and
oligomycin suggested the methodology could be manipulated in a predictable way.
91
2.27.2 Mitochondrial content
Mitochondrial content was assessed by TMRM staining of fibroblasts
seeded at 40 % confluency on 20 mm cover slips. Live cell TMRM images were
collected before (↓) and after (*↓) addition of 1 µM FCCP releasing the TMRM dye
into the cytosol to provide an image of the cell contour (Fig 2.4). Maximal intensity Z
projections of the TMRM signal were binarized in Image J and the proportion of the
binarized signal quantified relative to the cell contour (**↓). Quantification of the
binarized TMRM signal in Image J generated four values; minimum = 0, black image,
92
maximum = 255 white image, cell area and the mean TMRM intensity representing
the proportion of the highlighted area as white. The % TMRM density per area of cell
was calculated as [(mean TMRM intensity/255)*100].
2.27.3 GFP
SHSY5Y cells transiently transfected with GFP-UCP2 constructs (1 µg) were
seeded on 20 mm coverslips at 40 % confluency and assessed by confocal
microscopy 48 hours post transfection for the presence of green fluorescence.
Fluorescent images were acquired on an inverted laser scanning confocal
microscope (LSM510 Zeiss, Oberkochen, Germany) by exciting samples at 488 nm
with 10 % laser power directing speciments with the HFT 405/488/543 short pass
filter. Emitted red fluorescence was captured by 40x magnification through an oil
objective with numerical aperture of 1.2, a 515 nm long pass filter transmitted all
light above 515 nm. Pinhole diameter was set at 244 eliminating out of focus
fluorescence reaching the detector. The detector gain was adjusted to 688 allowing
for amplification of the fluorescent image and the amplifier offset set to 0.015 as the
93
cut off threshold below which pixels were displayed as black. Image resolution
parameters; pixel depth, scan speed and time course acquisition specifications are
described in Section 2.27.1.1.
2.28.1.1 Cellular respiration: oxygen electrode
Cellular oxygen utilization rates were analyzed polarographically using a
micro-Clark-type oxygen electrode (YSI, Hampshire, UK) [304]. The electrode was
calibrated with air saturated respiration buffer (HBSS) thermostatically maintained
at 37 °C. Whole fibroblasts or SHSY5Y cells (4x106) were resuspended in 250 µL of
respiration medium (HBSS) and extracellular oxygen levels monitored in a sealed
chamber over a period of 10 min. Uncoupled rates and mitochondrial proton leak
were measured in the presence of FCCP (0.5-2 µM) or oligomycin (1-4 mg/ml)
respectively. In experiments involving LRRK2 IN1, cells were treated with the
inhibitor under standard culturing conditions (1 µM, 90 min) with LRRK2 IN1
maintained in solution throughout the harvesting steps and respiration
measurements. Rates were calculated as the % decrease in the full scale deflection
(FSD) per minute with 0 % FSD determined by addition of 1 μM cyanide. The rates of
change of FSD were converted to changes in oxygen concentration assuming 406
nmol O per ml at the given temperature and corrected for protein content.
2.28.1.2 Method validation for the oxygen electrode
Extracellular oxygen concentrations measured by the oxygen electrode for
control fibroblasts showed a linear decrease in the signal over the course of the 10
minute measurement consistent with oxygen depletion as a result of cellular
respiration (Fig 2.5A). Oxygen levels were stable in the presence of the respiratory
chain complex IV inhibitor cyanide, reflecting inhibition of cellular oxygen
consumption and validating the specificity of our linear rate signals.
94
FCCP was used to investigate maximal respiratory capacity by the Clark type
oxygen electrode. Polarographic measurements of oxygen consumption in control
fibroblasts showed that the same concentration of FCCP (1 µM) used to dissipate the
mitochondrial membrane potential (Fig 2.3B) resulted in a two fold increase in the
rate of cellular oxygen consumption consistent with an uncoupling effect on the ETC
(Fig 2.5B). Titrating in more FCCP to 2 µM resulted in a reduced respiratory rate
indicative of mitochondrial toxicity as has previously been described for this
compound [305].
Oligomycin was used to investigate mitochondrial proton leak by the Clark
type oxygen electrode. Polarographic measurements of oxygen consumption in
control fibroblasts showed that addition of oligomycin at 4 mg/ml stabilized
extracellular oxygen levels (Fig 2.5C) consistent with inhibition of cellular respiration.
Lower oligomycin concentrations (1-3 mg/ml) showed a partial (70 %) decrease in
the linear rate of oxygen depletion rate signal potentially due to incomplete
inhibition of the mitochondrial ATPase. 4 mg/ml oligomycin were chosen for
subsequent experiments as the concentration required to inhibit the mitochondrial
ATPase in fibroblasts. Inhibition of mitochondrial respiration by cyanide and
oligomycin as well as mitochondrial uncoupling by FCCP suggests the methodology
could be manipulated in a predictable way.
95
2.28.2.1 Cellular respiration: phosphorescent analysis
Phosphorescent analysis of cellular oxygen utilization is an assay based on
the ability of oxygen to quench the excited state (lifetime) of the porphyrin based
extracellular MitoXpress probe. As cellular respiration occurs oxygen is depleted in
the surrounding environment correlating with an increase in the probe
phosphorescent signal.
Phosphorescent analysis of oxygen utilization required 50,000 fibroblasts
seeded per well on standard 96-well plates (Sarstedt) pre-coated with 0.007 %
collagen IV (Sigma) [306]. Measurements were conducted in 100 µl of air-
96
equilibrated DMEM supplemented with 1 mM pyruvate, 10 mM glucose (or 10 mM
galactose where stated), 20 mM HEPES (pH 7.4) and 100 nM of phosphorescent
MitoXpress-Xtra probe (Luxel Biosciences). Compounds used in the manipulation of
respiratory chain function (antimycin A (1 µM), rotenone (1 µM), FCCP and
oligomycin) were added at this point. Medium was overlaid with 50 µl of highly
viscous oil to prevent oxygen diffusion.
The plate was monitored at 37 °C on a Victor 2 time resolved fluorescent
(TR-F) reader (PerkinElmer, Cambridge, UK) with a Samarium filter set (340 nm
excitation, 642 nm emission). Duplicate TR-F intensity (F1, F2) readings were
recorded in a gated (100 µs) measurement window at two delay (t1=30, t2 =70 µs)
intervals. Measured TR-F intensity signals (F1, F2) for each well were converted into
phosphorescence life-time values as follows: τ = (t1–t2)/ln(F1/F2). To determine the
rate of change of probe phosphorescence in the presence of respiring cells lifetime
values were determined at 2 min intervals over 90 min. An initial temperature
equilibration phase of the probe lasting approximately 15 minutes was excluded
from the rate measurement.
Oxygen consumption rates were calculated as the rate of phosphorescent
life intensity change in relative phosphorescent units (RPU) by taking the slope of
the background subtracted linear signal. Rates were corrected for protein content.
2.28.2.2 Method validation of phosphorescent analysis
Phosphorescent analysis of control fibroblasts determined a linear increase
in the probe phosphorescent lifetime consistent with oxygen depletion in the
extracellular medium (Fig 2.6A). The rate of change of phosphorescent lifetime was
reduced to 30 % of the basal rate in untreated fibroblasts with the addition of
rotenone (1 µM) consistent with inhibition of mitochondrial respiration (Fig 2.6B).
97
Control fibroblasts were treated with increasing concentrations (0-5 µM) of
the chemical uncoupler. FCCP at 1 µM, a concentration shown to dissipate the
mitochondrial membrane potential resulted in a 50 % increase in phosphorescent
lifetime rate values with respiration rates reaching a maximum of 170 % at 1.5 µM
relative to untreated cells (Fig 2.6C). Rates were comparable to basal values at
higher FCCP concentrations. FCCP induced reduction in oxygen consumption at
higher concentrations are indicative of mitochondrial toxicity [305]. The
phosphorescent response to rotenone and FCCP suggested the methodology could
be manipulated in a predictable way.
98
2.29.1 Extracellular acidification
The pH-Xtra probe monitors cellular acid extrusion. The pH-Xtra
phosphorescence signal is modulated by pH changes such that increased
acidification causes increased phosphorescence. Extracellular acidification is an
indirect measure of the conversion of pyruvate to lactic acid in cells and provides
information on rates of glycolytic activity.
99
Rates of extracellular acidification were monitored as previously described
[305]. Fibroblasts (50, 000 cells per well) were seeded on standard 96- well plates
(Starstedt) pre-coated with 0.007 % collagen IV. Cells were maintained in buffer free
DMEM supplemented with 1 mM pyruvate, 10 mM glucose (or 10 mM galactose
where stated), 20 mM HEPES (pH 7.4) in a CO2 free incubator at 37 °C for 3 hours
prior to analysis. Medium was replaced with 100 µL of fresh solution containing 1
µM of pH-Xtra probe. Compounds used in the manipulation of respiratory chain
function (antimycin A (1 µM) and FCCP) were added at this point.
Time resolved fluorescence (TR-F) was recorded on a Victor 2 plate reader
thermostatically maintained at 37 °C utilizing a europium filter set (340 ± 35 nm
excitation, 615 ± 8.5 nm emission). Duplicate TR-F intensity (F1, F2) readings were
recorded in a gated (30 µs) measurement window at two delay (t 1=100, t2 =300 µs)
intervals allowing for calculations of phosphorescence lifetime values τ = (t1–
t2)/ln(F1/F2). For calculations of extracellular acidification τ was converted to pH =
(1893.4 - τ )/227.54 and [H+] = (ln(- pH)) (Fig 2.7A, B, C). To determine the rate of
change of probe phosphorescence, pH and [H+] lifetime values were acquired at 2
min intervals over 90 min. An initial temperature equilibration phase of the probe
lasting approximately 15 minutes was excluded from the rate measurement.
Extracellular acidification rates were calculated as the rate of change in [H+]
per minute by taking the slope of the background subtracted linear signal and
corrected for protein content.
2.29.2 Method validation for extracellular acidification measurements
Fluorescent analysis of the pH sensitive probe in the presence of control
fibroblasts determined a linear increase in the lifetime of probe phosphorescence
consistent acidification of the extracellular medium (Fig 2.7A).
100
FCCP effects on extracellular acidification rates showed a dose dependent
increase in the phosphorescent signal (Fig 2.7D) suggesting faster extracellular
acidification as a result of increased rates of anaerobic glycolysis. The titration
reached maximum rates, 50 % above basal acidification at 1 µM, comparable to
rates observed at higher FCCP concentrations consistent with glycolysis functioning
at maximum capacity despite mitochondrial toxicity.
Inhibition of the mitochondrial respiratory chain by antimycin A increased
the basal rate of extracellular acidification by 50 % (Fig 2.7E), mimicking the
response to FCCP treatment and consistent with upregulated glycolytic flux in
response to complex III inhibition. The probe phosphorescent lifetime response to
FCCP and antimycin A suggested the methodology could be manipulated in a
predictable way.
101
102
2.30 Cellular ATP content
Fibroblasts were seeded at a density of 10, 000 cells per well on opaquewalled 96 well plates (Nunc) and total cellular ATP was quantified in luminescence
units using CellTiter-Glo assay (Promega) according to manufacturer’s instructions.
Briefly, following removal of culture medium and a PBS wash, 100 µL of CellTiter-Glo
reagent containing renilla luciferase was added to each well. Cells were incubated
with the reagent (10 min, RT) allowing for cell lysis to occur. Plates were incubated
on a built in orbital shaker (Victor 2 plate reader) for 2 minutes following which end
point luminescence readings were acquired (excitation 375 nm, emission 495 nm).
Luminescence values were corrected for protein content.
2.31 Cellular ROS production: Array Scanner
Oxidised forms of dihydroethidium dye (DHE) have high affinity for DNA.
Upon oxidation, the fluorescent characteristics of reduced DHE (blue) change and
the red signal accumulates in the nucleus. Fluorescent properties of DHE can be
exploited to measure cellular ROS production.
Fibroblasts seeded at a density of 5000 cells/well on 96 well plates (Nunc)
were stained with 4’,6-diamidino-2-phenylindole (DAPI) in culture medium (1 µM, 37
°C, 20 min). Medium was replaced with HBSS containing 100 nM dihydroethidium
(Molecular probes). Generation of reactive oxygen species was monitored by
exciting samples with LED light source (543 nm (DHEox) and 361 nm (DAPI)) at 37 °C,
5% CO2 at 2 minute intervals for 1 hr (Array Scan VTI HCS Reader, Thermo Scientific,
Pittsburg, USA) [307]. Emitted Red fluorescence was captured within the Dapi
stained nuclear contour by a 20 X LD Plan-Neofluar objective through an automated
filter wheel monitoring the increase in red fluorescence over the course of the
measurement (Fig 2.8). Images were analyzed with the Cellomics Target Activation
103
Bio Application to assess the average DHE intensity per cell (n=100 per cell line) for
each time point allowing for rates of DHE oxidation to be quantified in real time.
104
Chapter 3: Results
Characterization of LRRK2 in cell models and brain tissue
Western blot analysis of LRRK2 expression has previously been used to detect
over-expressed LRRK2 in cultured cells as well as endogenous LRRK2 in cells of the
immune system and tissues like kidneys, lungs and lymph nodes. However, previous
reports using in house and commercial antibodies have struggled to detect low levels
of endogenous LRRK2, suggesting better antibodies or more sensitive detection
methods are required. Immunohistochemistry has been used to detect LRRK2 in
mouse and human brain tissue while immunoprecipitation has been reported to
detect endogenous brain LRRK2 protein and is often utilized for LRRK2 isolation from
cells for use in enzymatic activity assays, however its use for quantitative comparison
of LRRK2 expression has not been fully assessed. This chapter explored Western
blotting, BN PAGE, immunohistochemistry and a combined immunoprecipitation and
Western blot approach as methods to characterize LRRK2 protein expression in
SHSY5Y cells over-expressing wild type and G2019S LRRK2-V5, control and G2019S
PD patient fibroblasts, lymphoblasts and brain tissue, addressing the problem of
endogenous LRRK2 detection. Furthermore, the protein detection methods were
extended to characterization of LRRK2 cellular distribution and enzymatic activity
measurements to determine the properties of the wild type LRRK2 protein as well as
effects of the G2019S mutation on LRRK2 expression, localization and enzymatic
activity.
3.1. Western Blot analysis of LRRK2 protein expression in cell models and brain
tissue
To determine the sensitivity of LRRK2 detection by Western blot in our
endogenous and over-expressing LRRK2 cell model systems, four commercially
available antibodies (NT2, 100-500, C terminal, 3154-1) raised to different regions of
105
the LRRK2 protein (Fig 3.1A) were screened for their ability to detect LRRK2. The
polyclonal NT2 antibody was raised against a synthetic peptide encoding amino
acids 100-200 of the human LRRK2 sequence. The 100-500 monoclonal antibody was
raised against a synthetic peptide encoding amino acids 100-500 of the human
LRRK2 sequence. The C terminal monoclonal antibody was raised against a synthetic
peptide encoding amino acids 2079-2100 of human LRRK2 conjugated to a keyhole
limpet hemocyanin (KLH) sequence. The 3514-1 monoclonal antibody was raised
against a synthetic peptide encoding amino acids 970-2527 of the human LRRK2
sequence. All antibodies were raised in rabbits with target LRRK2 sequences 100 %
conserved in mouse and rat LRRK2 homologues. Antibodies were evaluated for their
ability to detect endogenous and over-expressed LRRK2 in human whole cell lysates
of untransfected SHSY5Y cells, fibroblasts, lymphoblasts, human brain cortex and
wild type LRRK2-V5 over-expressing SHSY5Y cells. Human cell lines c1 (fibroblasts)
and c2 (lymphoblasts) described in Table 2.2 were used for the analysis.
All four antibodies tested were able to detect a band at 280 kDa
corresponding to the full length protein in the LRRK2-V5 over-expressing SHSY5Y
cells as shown by the immunoreactive band in the (WT) lanes with the
corresponding signal absent in the untransfected cell line (Fig 3.1B). NT2 and 3514-1
antibodies produced an immunoreactive band of appropriate molecular weight for
full length LRRK2 in the lymphoblast cell line. None of the antibodies tested were
able to clearly detect 280 kDa immunoreactivity in fibroblast lysates (Fig 3.1B). Weak
280 kDa immunoreactivity was also observed in control SHSY5Y cells (Fig 3.1B) and
brain tissue for the 3514-1 but not 100-500 antibody (Fig 3.1C).
Additional immunoreactive species of lower molecular weight were
detected by NT2, 100-500 and 3514-1 in the over-expressing cell line (Fig 3.1B).
Some bands, 20-75 kDa, (i) were of equal intensity to those in untransfected control
106
cells while others, 100-150 kDa (ii) were present in the over-expressing cells only
suggesting species (ii) corresponded to the over-expressed LRRK2-V5 protein. Lower
molecular weight bands (30-100 kDa) were also detected in lymphoblast, fibroblast,
control SHSY5Y cells and human brain lysates (NT2, 100-500, 3514-1 (iii)) comigrating with those in the LRRK2-V5 over-expressing SHSY5Y cells potentially
representing endogenous forms of LRRK2. Lower molecular weight bands (30-150
kDa) absent in SHSY5Y cells were also detected in lymphoblast, fibroblast and
human brain lysates by the NT2 and 100-500 antibodies (iv), consistent with cell and
tissue specific antibody immunoreactivity (Fig 3.1B,C). As we did not have the
appropriate LRRK2 knockdown or knockout controls for fibroblast and lymphoblast
cell lines we were unable to rule out non-specific binding of antibodies in these cell
lines.
The lower molecular weight band patterns were unique to each antibody
potentially representing the position of antibody epitopes and the corresponding
LRRK2 fragment. In addition, variability was observed between blots as shown by
differences between the band pattern for wild type LRRK2 over-expressing SHSY5Y
cells, comparing 100-500 and 3514-1 immunoreactivities in (Fig. 3B and C),
complicating the identification of the true LRRK2 species. Our data suggests Western
blot detection of the LRRK2 protein can identify over-expressed full length LRRK2,
however, most antibodies identify additional immunoreactive bands of lower
molecular
weight
questioning
their
specificity.
Lower
molecular
weight
immunoreactive species were also observed in cells and tissues expressing
endogenous LRRK2 while immunoreactivity for the full length endogenous protein
was weak.
107
3.2 Characterization of LRRK2-V5 over-expressing SHSY5Y cells
3.2.1 Western blot analysis of LRRK2 protein in SHSY5Y cells
SHSY5Y cells stably expressing LRRK2-V5 were generated by J.M Cooper and
D. Chau, and positive clones expressing the wild type or G2019S LRRK2-V5 were
108
obtained. Western blot analysis of untransfected, wild type or G2019S LRRK2-V5
over-expressing SHSY5Y whole cell lysates revealed a single 280 kDa band
immunoreactive for V5 in SHSY5Y cells transfected with wild type and G2019S
LRRK2-V5 absent in control SHSY5Y cell extracts (Fig 3.2A). The band intensity in wild
type extracts was 2 fold higher relative to the G2019S clone when corrected for
protein loading by GAPDH immunoreactivity. To ensure consistent expression of the
LRRK2 constructs, protein levels were assessed with increasing passage number.
Positive V5 immunoreactivity was detected by Western blot for both wild type (Fig
3.2B) and G2019S (Fig 3.2C) clones for passages 7-22. LRRK2 expression was reduced
by 40 % in G2019S LRRK2-V5 over-expressing cells at passages 21 and 22 when
corrected for protein loading by GAPDH immunoreactivity suggesting the stability of
over-expressed G2019S LRRK2 protein was affected at these passages. All functional
analyses were carried out for passages 8-20 to compare the effects of overexpressed wild type and G2019S LRRK2.
3.2.2 LRRK2 mutation screen in SHSY5Y cells
To determine whether the LRRK2 pathogenic G2019S mutation was
expressed in wild type and G2019S LRRK2-V5 over-expressing cell lines, mRNA was
extracted from each clone, reverse transcribed and the corresponding coding DNA
analysed for the presence of the G2019S mutation in exon 41 of the LRRK2 gene.
Successful amplification of the 700 base pair fragment from exon 41 of LRRK2 cDNA
in SHSY5Y cells was confirmed by the presence of a single DNA band co-migrating
with a fragment amplified from the LRRK2 plasmid absent in the water blank
reaction (Appendix Fig A4). DNA sequencing of wild type PCR products confirmed
100 % base pair complementarity to the human LRRK2 gene registered on the
UniProt database (Fig 3.2D). cDNA from SHSY5Y cells stably over-expressing mutant
LRRK2 protein was identified to have the adenine nucleotide at base pair position
109
6055 with endogenous wild type LRRK2 (guanine at position 6055) not resolved by
sequencing potentially due to the differential level of expression (Fig 3.2E). The
guanine to adenine substitution results in a codon encoding the serine (S) amino
acid, confirming expression mutant LRRK2 protein in SHSY5Y cells.
110
3.2.3 Western blot analysis of LRRK2 sub-cellular distribution in SHSY5Y cells
Having demonstrated wild type and G2019S LRRK2-V5 over-expression in
SHSY5Y cells, LRRK2 cellular distribution was evaluated in these cells lines. SHSY5Y
cells over-expressing LRRK2 were ruptured by nitrogen cavitation and fractionated
by differential centrifugation. Western blot analysis of wild type LRRK2-V5
expressing cell fractions identified V5 immunoreactivity in all the isolated SHSY5Y
compartments with highest abundance (47 %) in the 250,000 g and cytosolic
fractions (Fig 3.3A). Quantification of LRRK2 immunoreactive bands by densitometry
approximated 5 % (n=1) of cellular LRRK2 in each of the 10-47,000 g fractions.
Compartmental characterization identified lysosomal abundance in the 47,000 g
fraction as determined by the double immunoreactive GCase band of approximately
60 kDa representing the glycosylated and non glycosylated form of the lysosomal
enzyme. Mitochondrial enrichment was found in the 47,000 g (60 %), 27,000 g (30
%) and 1,000 g (10 %) fractions, as determined by immunoreactivity to the Core
Complex III mitochondrial marker. Immunoreactivity for the endoplasmic reticulum
marker, calreticulin, was detected in 27,000 g (55 %), 12,000 g (40 %) and 47,000 g
(5 %) fractions. Synaptophysin immunoreactivity, representing synaptic vesicle
abundance were identified in 250,000 g (35 %) and cytosolic (55 %) compartments
with the remaining 10 % synaptophysin immunoreactivity distributed equally
between fractions 1-45,000 g. These data shows that wild type, over-expressed
LRRK2-V5 can be found abundant in the cytosolic and vesicle enriched cellular
compartments, with lesser abundance in fractions containing ER, lysosomal and
mitochondrial markers.
To determine whether G2019S expression is associated with altered LRRK2
cellular distribution in G2019S LRRK2-V5 over-expressing SHSY5Y cells, differential
centrifugation and Western blot analysis was performed on the mutant cell line.
111
G2019S LRRK2 compartmental distribution in SHSY5Y cells reflected that of the wild
type protein with highest LRRK2 abundance identified in 250,000 g (60 %) and
cytosolic (30 %) fractions with 6 % (n=1) of the total V5 immunoreactivity distributed
equally between the fractions 10-47,000 g (Fig 3.3B). Unlike wild type overexpressed LRRK2, 3 % of G2019S LRRK2 was identified in the 1,000 g fraction,
corresponding to unbroken cells and nuclear components, potentially representing
incomplete cellular disruption. Compartmental characterization of G2019S LRRK2
expressing fractions identified lysosomal abundance (95 %) in the 12,000 g fraction
as determined by the double immunoreactive GCase band of approximately 60 kDa.
Mitochondrial enrichment was found in the 47,000 g (30 %), 27,000 g (60 %) and
1,000 g (10 %) fractions, as determined by immunoreactivity to the Core Complex III
mitochondrial marker. Calreticulin immunoreactivity, representing the endoplasmic
reticulum was abundant in 27,000 g (55 %) and 12,000 g (35 %) fractions, while
synaptophysin immunoreactivity was identified in 250,000 g (60 %) and cytosolic (35
%) fractions as well as 5 % of the total synaptophysin immunoreactivity distributed
equally between fractions 1-47,000 g.
Lysosomal and mitochondrial markers in fractionated G2019S LRRK2-V5
over-expressing SHSY5Y cells were detected in fractions isolated at lower
centrifugation speeds (12,000 g) relative to the wild type LRRK2-V5 over-expressing
SHSY5Y cells (47,000 g) potentially reflecting the smaller size of mitochondrial and
lysosomal organelles. ER and synaptic markers were identified in fractions isolated
by comparable centrifugation speeds (12,000 and 27,000 g) and (250,000 g
respectively) for both wild type and G2019S LRRK2-V5 over-expressing cells. Our
data shows that over-expressed LRRK2 V5 can be found abundant in the cytosolic
and vesicle enriched cellular compartments, with smaller amounts present in
fractions containing ER, lysosomal and mitochondrial markers. In addition G2019S
112
LRRK2-V5 cellular localization mirrors the cellular distribution of the wild type
LRRK2-V5 protein in SHSY5Y cells. However, the difference in centrifugation speeds
required to isolate the majority of the mitochondrial and lysosomal pools may
reflect a difference in organelle boyant density between the wild type and G2019S
LRRK2-V5 over-expressing SHSY5Y cells.
3.2.4 Immunocytochemistry of LRRK2 sub-cellular distribution in SHSY5Y cells
In addition to Western blot analysis, wild type LRRK2-V5 over-expression
and cellular distribution in SHSY5Y cells was characterized by immunocytochemistry.
In an attempt to visualise LRRK2 over-expression by imunocytochemistry four
permeabilization and fixation conditions were tested; methanol and acetone,
paraformaldehyde
and
Triton
X100,
paraformaldehyde
and
methanol,
paraformaldehyde and sodium citrate. Immunocytochemistry utilizing methanol and
113
acetone conditions confirmed LRRK2-V5 over-expression in wild type expressing
SHSY5Y cells as determined by V5 immunoreactivity absent in untransfected cells
(Fig 3.4A). Immunostaining conditions utilizing paraformaldehyde in combination
with Triton X100 and methanol permeabilization or boiling in sodium citrate
produced V5 immunoreactive signals in control SHSY5Y cells of comparable
intensities to that observed in wild type LRRK2 over-expressing SHSY5Y cells. V5
positive immunoreactivity in paraformaldehyde fixed control SHSY5Y cells was likely
due to non specific binding of the antibody under these conditions.
Methanol acetone processing of SHSY5Y cells identified V5 positive
immunoreactivity in wild type LRRK2 over-expressing SHSY5Y cells as spread
diffusely throughout the cell body with some immunoreactive punctate structures
detected (Fig 3.4B). Co-localization with compartmental markers was not assessed
due to this diffuse nature of the LRRK2 signal. G2019S LRRK2-V5 over-expression in
SHSY5Y cells could not reliably be identified by immunocytochemistry this may
reflect the 2 fold lower expression of the G2019S LRRK2-V5 protein relative to wild
type LRRK2-V5 over-expressing cells, questioning the sensitivity of this detection
method.
114
3.3 Characterization of endogenous LRRK2 in lymphoblasts
3.3.1 Western blot analysis of LRRK2 protein in lymphoblasts
Western blot analysis of LRRK2 protein detected endogenous expression in
lymphoblasts with the NT2 antibody (Fig 3.1B) consistent with high LRRK2
abundance as previously reported for this cell type [244]. Lymphoblast cells are
115
derived from peripheral blood mononuclear cells (PMBCs) isolated from whole
blood. To determine whether LRRK2 expression in control lymphoblasts reflects
LRRK2 protein abundance in PMBCs, LRRK2 expression was compared by Western
blot analysis. It was not possible to obtain PMBCs from any of the lymphoblast
donors and hence PMBCs from 5 healthy control subjects of different ages (23-62
years) were screened for LRRK2 expression. A lymphoblast cell line (control subject,
56 years) was used for comparison. Positive NT2 immunoreactivity in the 280 kDa
region confirmed LRRK2 expression in PMBCs with LRRK2 levels comparable
between transformed and primary cells relative to GAPDH immunoreactivity (Fig
3.5A). LRRK2 PMBC expression was similar between the five control donors
suggesting LRRK2 lymphoblast expression reflects levels in untransformed cells.
PMBCs purified by density gradient isolation consist of a combination of T
and B lymphocytes of differential abundance [308] while lymphoblasts are derived
primarily from Epstein-Barr virus transformed B cells. To determine which
lymphocyte subpopulations express LRRK2 protein PMBC cells (control subject, 49
years) were used for affinity purification and characterization of T (CD4, CD8) and B
(CD19) lymphocyte cell subtypes. PMBCs were fractionated by positive selection
using magnetically tagged antibodies specific for CD4, CD8 and CD19 lymphocytes
and characterised for LRRK2 protein expression by Western blotting. Western Blot
analysis of purified T and B cell extracts produced a strong NT2 immunoreactive
band in the CD19 cells, 10 fold lower intensity immunoreactive signal in CD8 relative
to GAPDH when corrected for protein loading but no clear LRRK2 immunoreactivity
in CD4 cells (Fig 3.5B) consistent with LRRK2 protein abundance in B cells and
substantially lower expression in T cells.
Having shown that LRRK2 protein expression in transformed lymphoblasts
reflects the protein levels of CD19 positive B cells from which the lymphoblasts were
116
derived, we set out to determine whether LRRK2 protein abundance is altered with
G2019S expression. LRRK2 levels were compared between control and G2019S
expressing lymphoblasts. Western blot analysis of control and G2019S lymphoblast
cell lines was carried out by Susannah Horan (Clinical Neurosciences, UCL). As the
NT2 antibody was not available at the time, the previously validated rabbit
polyclonal NB300-268 was used (refer to diagram in 3.1A for epitope details).
Western blot analysis of NB300-268 immunoreactivity indicated variability in LRRK2
levels between the cell lines within the control and G2019S lymphoblast groups (Fig
3.5C). When immunoreactivities were quantified, variability of 10 fold intensity
difference was noted between highest and lowest LRRK2 expressing lymphoblast
lines. The variability of LRRK2 expression in lymphoblasts complicated the
assessment of G2019S expression effects on LRRK2 abundance.
117
3.3.2 Western blot analysis of LRRK2 sub-cellular distribution in lymphoblasts
Western blot analysis of LRRK2 protein expression with the NT2 antibody
determined that LRRK2 levels were approximately 3 fold lower in a high LRRK2
expressing lymphoblast cell line (c2) relative to wild type LRRK2-V5 over-expression
118
in SHSY5Y cells (Fig 3.1B). To compare the cellular distribution of endogenous LRRK2
in lymphoblasts to the data obtained for wild type and G2019S LRRK2-V5 SHSY5Y
cells, differential centrifugation and Western blot analysis were carried out on the c2
lymphoblast cell line (Fig 3.6). 280 kDa NT2 immunoreactivity was abundant in the
cytosolic (60 %) and 250,000 g (35 %) compartments with 5% of the total NT2
immunoreactivity distributed equally between 1-47,000 g compartments (Fig 3.6),
comparable to the V5 immunoreactive pattern observed for wild type and G2019S
LRRK2-V5 expressing SHSY5Y cells (Fig 3.3A, B). Higher molecular weight NT2
immunoreactive species (approximately 290 and 300 kDa) were detected in 12,000 g
and 250,000 g fractions, absent in the 1,000, 27,000, 47,000 g and cytosolic
fractions.
Characterization of cellular compartmental distribution identified lysosomal
abundance in the 12,000 g (60 %), 27,000 g (30 %) and 1,000 g (10 %) fractions as
determined by GCase immunoreactivity. A single immunoreactive band was
detected by the GCase antibody in these fractions suggesting lymphoblast derived
GCase exists in a single 60 kDa form, potentially corresponding to the unglycosylated
enzyme. Mitochondrial abundance was identified in 1,000 g (45 %) and 12,000 g (50
%) fractions with additional bands in 27,000 g and 47,000 g (5 %) compartments, as
determined by Core CIII immunoreactivity. Calreticulin (ER) markers were present in
the 47,000 g (40 %) fractions where a double band was detected potentially
representing a post translationally modified form of calreticulin [309]. A single band
(*) immunoreactive for the calreticulin antibody was identified in the 27,000 g
fraction accounting for 60 % of the total calreticulin immunoreactivity, potentially
representing the unglycosylated form of the protein.
Similar to the results obtained in the LRRK2-V5 over-expressing SHSY5Y
system, endogenous LRRK2 in lymphoblasts can be found abundant in the cytosolic
119
compartment, as well as minor levels in fractions containing ER, lysosomal and
mitochondrial markers. These results suggest that LRRK2 cellular distribution was
not affected by LRRK2 over-expression.
3.4 LRRK2 immunoprecipitation
3.4.1 Characterization of LRRK2 immunoprecipitation for quantitative comparison
We addressed the problem of endogenous LRRK2 detection in fibroblasts,
SHSY5Y cells and brain tissue. The immunoprecipitation protocol, based on
methodology described by Alessi et al, was extended to the detection of
endogenous LRRK2 protein [215]. Additional characterization of methodology was
carried out to validate the use of this method for quantitative comparison.
LRRK2 immunoprecipitated from control and LRRK2 knockout mouse brain
cortex or SHSY5Y cells over expressing wild type LRRK2-V5 was assessed by Western
blot analysis. Successful immunoprecipitation of full length LRRK2 from mouse
120
cortex was confirmed by the presence of a 280 kDa band undetected in the
corresponding LRRK2 knockout brain immunoprecipitates and the no antibody
control lane as determined by Western blot analysis with 3514-1 antibody (Fig 3.7A).
LRRK2 immunoreactive bands co-migrated with full length wild type LRRK2-V5 in
whole cell extracts and immunoprecipitates from SHSY5Y over-expressing cells. A
250 kDa species (*) was also present in immunoprecipitates of wild type LRRK2-V5
over-expressing SHSY5Y cells but absent in mouse brain and no antibody control
immunoprecipitates
suggesting
the
presence
of
an
additional
LRRK2
immunoreactive species in this cell type (Fig 3.7A).
The solubility of LRRK2 in buffer used for immunoprecipitation was assessed
in SHSY5Y cells over-expressing LRRK2 by Western blot analysis of detergent soluble
and insoluble fractions. Non ionic detergent (1 % Triton X100 and NP-40) containing
buffers retained little of the cellular LRRK2 in the insoluble pellet with >95 % of
LRRK2 found in the soluble fraction as determined by V5 immunoreactivity (Fig
3.7B). Substitution of non ionic with ionic (2 % CHAPS, 1 % Cholate) or zwitterionic (1
% DDM) detergents recovered comparable levels (>95 %) of LRRK2 in the soluble
faction. An additional band immunoreactive for the V5 antibody was detected in all
soluble fractions (*). The relative size of the band (250 kDa) was comparable to the
species
identified
in
the
LRRK2-V5
over-expressing
SHSY5Y
cells
by
immunoprecipitation (Fig 3.7A). Our data suggests >95 % of the cellular LRRK2 pool
can be recovered in IP buffer supplemented with ionic, non ionic and zwitter ionic
detergents in LRRK2 over expressing SHSY5Y cells.
The linearity of the IP signal was assessed by immunoprecipitation of
endogenous LRRK2 from increasing concentrations (100-600 µg) of fibroblast lysate
extracted in 1 % Triton X100 containing IP buffer. 3514-1 immunoreactivity was
detected in all fibroblast immunoprecipitates (Fig 3.7C) consistent with the presence
121
of LRRK2 protein in fibroblasts. The lower molecular weight species detected by IP
and Western blot in wild type LRRK2-V5 over-expressing SHSY5Y cells (Fig 3.7A) was
absent in the fibroblast IP. The 3514-1 band intensity was determined to be in linear
proportion to the concentration of fibroblast input lysate as shown by plotting the
band intensity relative to µg of lysate (Fig 3.7D).
Our data demonstrated the specificity of the LRRK2 immunoprecipitation
procedure as determined by analysis of wild type and LRRK2 knockout brain tissue,
efficient LRRK2 solubilization (>95 %) using the IP buffer conditions in LRRK2 over
expressing SHSY5Y cells and signal linearity based on the IP LRRK2 recovery from
fibroblast input lysates confirming the suitability of IP for quantitative comparison of
LRRK2 protein expression.
122
3.4.2 LRRK2 kinase activity in lymphoblasts and SHSY5Y cells
Immunoprecipitation of LRRK2 from lymphoblasts and LRRK2 overexpressing HEK293 cells has previously been shown to produce a LRRK2 protein with
sufficient purity for enzymatic activity measurements in vitro [202, 215]. LRRK2 was
immunoprecipitated from 1 mg of 1 % Triton X100 lysates of control lymphoblasts,
123
wild type and G2019S LRRK2-V5 over-expressing SHSY5Y cells and assessed for
kinase activity. Six lines of control lymphoblasts were used to determine the
potential correlation between LRRK2 expression and kinase activity. Successful
immunoprecipitation of LRRK2 was determined by the single 280 kDa 3514-1
immunoreactive band identified for all samples. The intensity of 3514-1
immunoreactivity was found to be variable for the lymphoblastoid cells (Fig 3.8A),
consistent with the variable LRRK2 protein expression identified by Western blot
analysis of whole cell lysates (Fig 3.5C). The 3514-1 band intensity was stronger for
immunoprecipitates of wild type and G2019S LRRK2-V5 SHSY5Y cells consistent with
greater LRRK2 protein expression in this cell type.
Following immunoprecipitation, LRRK2 isolated from control lymphoblasts
and over-expressing SHSY5Y cells was assessed for kinase activity by rate
measuments of radiolabelled 32P incorporation into the LRRKtide substrate. Kinase
activity measurements were carried out in Dundee with the help of Fransisco InestaVaquera. GST-LRRK2 isolated from HEK293 cells was used as a positive control for
the reaction [202]. LRRKtide phosphorylation by the purified GST-LRRK2 was
identified by detection of incorporated radioactivity as counts per minute abolished
in the presence of a LRRK2 kinase inhibitor LRRK2-IN1, suggesting GST-LRRK2 was
active and confirming LRRKtide as a GST-LRRK2 substrate (Fig 3.8B).
Calculating specific activities from cpm values of the LRRK2 enzyme
immunoprecipitated from lymphoblasts determined substantial variability between
the control lymphoblast cell lines, with values ranging from 0.6-1 µmol/min/mg
protein (Fig 3.8C). Due to variability in LRRK2 specific activity for control
lymphoblasts, LRRK2 kinase activity was not assessed in G2019S expressing
lymphoblast cells.
124
The specific activity of wild type and G2019S LRRK2-V5 immunoprecipitated
from SHSY5Y over-expressing cells was 4-8 fold lower relative to the lymphoblast
cells (Fig 3.8C) suggesting the over-expressed LRRK2 kinase is potentially less active
in the SHSY5Y cells. The specific activity of immunoprecipitated G2019S LRRK2-V5
was approximately 2 fold greater relative to the wild type LRRK2 consistent with
higher kinase activity associated with G2019S LRRK2-V5 (Fig 3.8C).
Our data show that the active LRRK2 enzyme can be effectively
immunoprecipitated from LRRK2-V5 over-expressing SHSY5Y cells and lymphoblasts.
In addition, we have demonstrated that the kinase activity of LRRK2 is greater for
LRRK2 immunoprecipitated from lymphoblasts relative to LRRK2-V5 over-expressed
in SHSY5Y cells while the G2019S LRRK2 mutation correlates with the characteristic
increase in the kinase activity of the mutant protein in our SHSY5Y LRRK2-V5 overexpressing system.
125
3.5 Characterization of fibroblast cell lines
3.5.1 Immunoprecipitation of LRRK2 protein from fibroblasts
Six control and eight G2019S patient fibroblast cell lines were generated
with details about the age of donor and age at collection described in Table 2.2.
Having demonstrated endogenous LRRK2 protein expression in fibroblasts (Fig 3.7C)
and evaluated the IP procedure for use in quantitative comparison of LRRK2 protein
expression, we set out to determine whether the G2019S mutation correlated with
126
altered LRRK2 protein abundance in patient derived fibroblasts. Endogenous LRRK2
was immunoprecipitated from equivalent amounts of input Triton X100 lysates (1
mg) of 3 control and 3 G2019S patient fibroblast cell lines. The relative band
intensity of the LRRK2 immunoreactive signal was comparable for wild type and
G2019S fibroblast cell lines when corrected for IP input by quantifying the GAPDH
immunoreactivity (Fig 3.9A). Unlike control and G2019S lymphoblasts (Fig 3.5C),
LRRK2 protein expression in control and G2019S patient fibroblasts was less
variable, despite differences in the age at sample collection of the donors.
3.5.2 LRRK2 mutation screen in control and G2019S fibroblasts
To confirm expression of the pathogenic G2019S mutation, mRNA was
extracted from each fibroblast cell line (6 control, 8 G2019S), reverse transcribed
and the corresponding coding DNA analysed for the presence of the G2019S
mutation on exon 41 of the LRRK2 gene. Successful amplification of the 700 base
pair fragment from exon 41 of LRRK2 cDNA in fibroblasts confirmed by the presence
of a single DNA band co-migrating with a fragment amplified from the LRRK2
plasmid absent in the water blank reaction (Appendix Fig A3). DNA sequencing of
wild type PCR products confirmed 100 % base pair complementarity to the human
LRRK2 gene registered on the UniProt database for all control cell lines (Fig 3.9B). All
clinically diagnosed PD G2019S cases were found to contain the heterozygous
mutant allele as determined by 50 % adenine abundance at base pair position 6055
of the LRRK2 gene (Fig 3.9C).
3.5.3 LRRK2 mRNA expression in control and G2019S fibroblasts
To determine whether LRRK2 expression was altered in cells expressing the
G2019S mutation, quantitative real time PCR analysis was carried out on 6 control
and 8 G2019S fibroblast cell lines. Amplification of LRRK2 cDNA produced a PCR
product in cycles 22-24 (equivalent GAPDH in cycles 14-16) of the RT PCR reaction
127
indicative of low levels of endogenous LRRK2 expression (Fig 3.9D). Quantification of
LRRK2 cycle threshold relative to the cDNA loading control marker GAPDH for
control versus G2019S fibroblast lines revealed a similar level of LRRK2 mRNA
abundance (Fig 3.9E), consistent with equivalent LRRK2 mRNA expression in wild
type and G2019S LRRK2 fibroblasts. Our data is consistent with comparable LRRK2
mRNA and protein abundance in control and G2019S patient fibroblasts.
128
3.6 Analysis of LRRK2 knockdown in SHSY5Y cells
In addition to the quantitative comparison of LRRK2 protein expression
between control and G2019S patient fibroblasts, LRRK2 protein detection by the
combined immunoprecipitation and Western blot approach was extended to the
129
characterization of the LRRK2 siRNA knockdown model in SHSY5Y cells. LRRK2 siRNA
was introduced into control SHSY5Y cells by the method of transient transfection.
For prolonged LRRK2 gene silencing, cells were re-transfected with siRNA at day 3.
SHSY5Y cells were analysed after 3 and 6 days of silencing for LRRK2 mRNA
expression by quantitative real time PCR and LRRK2 protein abundance by
immunoprecipitation and Western blot analysis.
Following siRNA treatment, mRNA was extracted from SHSY5Y cells, reverse
transcribed to cDNA and LRRK2 expression quantified by real time PCR. Data was
expressed as % LRRK2 expression relative to untreated cells. Quantification of LRRK2
amplification relative to GAPDH CT values identified a 70 % reduction in LRRK2
mRNA expression in SHSY5Y cells treated with siRNA for 3 and 6 days when
compared to untreated cells (Fig 3.10A) consistent with a reduction in LRRK2 gene
expression.
LRRK2 was immunoprecipitated from 1mg of 1 % Triton X100 lysates of
control SHSY5Y cells and cells treated with LRRK2 siRNA. Western blot analysis of
immunoprecipitates revealed a 65 % reduction in 3514-1 280 kDa immunoreactivity
3 days post transfection relative to untreated cells when corrected for GAPDH
immunoreactivity representing IP input (Fig 3.10B, C). The observed 65 % reduction
in 3514-1 band intensity was consistent with reduced LRRK2 protein expression in
LRRK2 siRNA treated SHSY5Y cells. Extending the LRRK2 siRNA treatment to 6 days
reduced the 3514-1 immunoreactive LRRK2 band to 85 % of the untreated SHSY5Y
cells suggesting SHSY5Y LRRK2 protein levels were less abundant following
prolonged incubations with LRRK2 siRNA. An additional 3514-1 immunoreactive
band at 250 kDa (*) was identified for endogenous LRRK2 immunoprecipitated from
SHSY5Y cells with the approximate molecular weight comparable to the species
observed in the IP of the wild type LRRK2-V5 over-expressing SHSY5Y cells (Fig 3.7A).
130
The intensity of this band was reduced following LRRK2 siRNA treatment suggesting
the band represents LRRK2 protein.
Analysis of the LRRK2 siRNA SHSY5Y knockdown model not only revealed an
efficient reduction of endogenous LRRK2 protein level but also further validated the
specificity of the IP procedure as the intensity of the 280 kDa LRRK2 immunoreactive
bands was sensitive to LRRK2 siRNA.
131
3.7 Immunoprecipitation of endogenous LRRK2 protein from brain
Using the combined immunoprecipitation and Western blot protocol we
analysed LRRK2 expression in two control human brains of 24 hour post mortem
delay. LRRK2 was immunoprecipitated from equivalent amounts of Triton X100
lysates (1 mg) of human cerebellum, caudate nucleus, medial putamen and A10
region of the cortex. Western blot analysis revealed a weak 3514-1 280 kDa
immunoreactive band in cerebellum 2 immunoprecipitate (←), co-migrating with
the 280 kDa immunoreactive species detected in the fibroblast IP. This band was
only seen following prolonged blot exposure, potentially representing low abundant
full length LRRK2 protein. LRRK2 immunoreactivity was not readily detected in
human brain caudate nucleus, medial putamen or cortical immunoprecipitates using
this approach (Fig 3.11A).
LRRK2 was immunoprecipitated from equivalent amounts of Triton X100
lysates (1 mg) of control human cortex and amygdala brain regions with post
mortem delay times of 3h30-8h30 as well as an idiopathic PD brain of 1h30 and
Alzheimer’s disease brain of 1h45 post mortem delay times respectively. We were
unable to obtain control brains of 1h30 and 1h45 post mortem delays. Western blot
analysis of LRRK2 immunoprecipitates identified weak 3514-1 280 kDa
immunoreactivity for the 1h30 cortical extract of an idiopathic PD case (←). No clear
bands were detected in the other cortical samples or amygdala immunoprecipitates
suggesting low LRRK2 human brain abundance or a problem with the
immunoprecipitation procedure (Fig 3.11B).
132
As LRRK2 could be immunoprecipitated from fresh mouse brain cortical
extracts (Fig 3.7A), we assessed LRRK2 brain distribution in different regions of the
mouse brain. Immunoprecipitation and Western blot analysis of LRRK2 from brain
stem, olfactory lobe, mid brain, striatum, cortex and cerebellum regions of the
mouse brains revealed a 3514-1 immunoreactive 280 kDa band in all brain regions
analysed (Fig. 3.12A). This 280 kDa band co-migrated with the 3514-1
immunoreactivity of LRRK2 fibroblast immunoprecipitates and was absent in the IP
where the LRRK2 antibody was omitted. LRRK2 immunoprecipitation from mouse
brain produced strongest immunoreactivity in the striatum, two fold greater than
that of the cerebellum and cortex, 5 fold higher when compared to the band
intensity in the brain stem and 10 fold greater than that of the 280 kDa band in the
midbrain and olfactory lobe (Fig 3.12A). Comparable 3514-1immunoreactive
patterns were detected in the equivalent regions of the rat brain (Fig 3.12B) with
133
striatal band intensity 10 fold greater when compared to that detected in the
midbrain. 3514-1 immunoreactivity was 2 fold greater in the cortex when compared
to the striatum with a higher molecular weight band (←) consistently detected in
immunoprecipitates from this region but absent in the equivalent mouse brain
areas. 280 kDa immunoreactivities could not be detected in the rat olfactory lobe or
brain stem at these exposure times potentially due to lower LRRK2 protein
abundance. Extending these analyses to non human primates, LRRK2 was
immunoprecipitated from equivalent regions of MPTP treated marmoset brains. We
were unable to obtain brains from untreated control animals. Western blot analysis
of LRRK2 immunoprecipitates produced a single 280 kDa 3514-1 immunoreactive
band in the striatum and a band of 3 fold greater intensity in the cortex (Fig 3.12C).
Weaker (10 fold) 3514-1 immunoreactivity was detected in the marmoset brain
stem, olfactory lobe and midbrain, reflecting the LRRK2 protein distribution pattern
identified in the equivalent brain regions of the rodent brains. Although the pattern
of LRRK2 protein distribution is comparable between the mouse, rat and marmoset
brains, we cannot rule out the potential effects of MPTP on LRRK2 expression.
Our data shows that endogenous LRRK2 can be readily immunoprecipitated
from rodent and non human primate brain tissue in contrast to the human brain
tissue raising the question whether the contents of the human brain extracts could
potentially be interfering with the IP procedure. To test this fibroblast IP lysates
were spiked with equal concentrations of mouse or human brain cortex IP lysates
and subject to LRRK2 immunoprecipitation. LRRK2 readily immunoprecipitated from
fibroblast and mouse cortical extracts as determined by 3514-1 280 kDa
immunoreactivity
(Fig 3.12D). The
relative
band
intensity
in
fibroblast
immunoprecipitates spiked with a mouse brain extract was equivalent to the sum of
the band intensities detected in the individual fibroblast and mouse brain samples.
134
No clear immunoreactivity was detected for the LRRK2 immunoprecipitates from a
human cortical extract while a 3 fold reduction in signal intensity was observed for
the fibroblast immunoprecipitate spiked with the human lysate relative to the band
intensity detected in the pure fibroblast immunoprecipitate (Fig 3.12D) suggesting
the composition of the human brain lysate may be affecting immunoprecipitation.
135
3.8 Comparison of LRRK2 expression in cell models and brain tissue
Due to potential interference in the human LRRK2 IP, mouse brains were
used for comparative analysis of LRRK2 protein expression with fibroblasts,
lymphoblasts, control and LRRK2-V5 over-expressing SHSY5Y cells. LRRK2
immunoprecipitated from equal concentrations (1 mg) of input lysates confirmed
highest endogenous LRRK2 abundance in lymphoblasts. Levels were 10 fold lower in
fibroblasts and 50 fold lower in control SHSY5Y cells (Fig 3.13) based on
quantification of 3514-1 280 kDa immunoreactivity. In mouse brains, 3514-1
immunoreactivity was greatest in the striatum and cortex, consistent with previous
findings (Fig 3.12A) comparable to the relative band intensities detected in SHSY5Y
cells and 7 fold lower than those in fibroblast immunoprecipitates. Wild type LRRK2V5 over-expression in SHSY5Y cells was quantified as 3.6 fold higher than in control
lymphoblast
and
approximately
25
fold
greater
than
in
fibroblast
immunoprecipitates as determined by 3514-1 280 kDa immunoreactivity (Fig 3.13).
Additional lower molecular bands of 250 kDa were identified in over-expressing cells
and lymphoblasts (←) while longer exposures revealed equivalent immunoreactive
bands in fibroblast IPs (←) resembling the bands previously identified by Western
blotting in LRRK2-V5 over-expressing SHSY5Y cells (Fig 3.7B). These species were also
identified in control SHSY5Y cells and reduced upon LRRK2 siRNA treatment (Fig
3.10B, C) suggesting the 250 kDa species is found in both the LRRK2 over-expressing
cell lines and systems expressing endogenous LRRK2.
Our results show that control SHSY5Y cells and fibroblasts express LRRK2
protein at levels similar to those detected in the brain while the LRRK2 overexpressing SHSY5Y model has comparable LRRK2 expression relative to endogenous
LRRK2 abundance in lymphoblasts.
136
3.9 Analysis of LRRK2 protein expression by blue native PAGE electrophoresis
As an alternative analysis of LRRK2 in the human brain samples blue native
PAGE combined with Western blotting was carried out with the 3514-1 LRRK2
antibody. To determine whether preserving the native structure of LRRK2 would
improve the 3514-1 antibody specificity, blue native PAGE electrophoresis was
carried out on soluble cell extracts of SHSY5Y cells over-expressing wild type LRRK2V5. In parallel, blots were probed with the V5 antibody to compare the specificities
of the two antibodies. Cells were solubilised in extraction buffer in the absence and
presence of detergent (1 % DDM or Triton X100) to determine whether inclusion of
detergent would improve LRRK2 recovery. A single V5 280 kDa immunoreactive
band (i) was detected by Western blot in soluble extracts of LRRK2-V5 overexpressing SHSY5Y cells. A 3514-1 280 kDa immunoreactive band of comparable size
and signal intensity was also identified (Fig 3.14A) suggesting the two antibodies
recognised the same species. Longer exposures revealed 900 kDa immunoreactive
137
species for both the V5 and 3514-1 antibodies (ii), potentially corresponding to a less
abundant higher molecular weight wild type LRRK2-V5.
Solubilisation of wild type LRRK2-V5 over-expressing SHSY5Y cells in buffer
containing 1 % DDM, revealed a poorly resolved V5 immunoreactive smear by
Western blot with 280 kDa (i) and 900 kDa (ii) bands visible at short exposures.
Solubilisation of wild type LRRK2-V5 over-expressing SHSY5Y cells in buffer
containing 1 % Triton X100 also produced some V5 immunoreactive smearing and a
280 kDa band at low exposure (i) with a weak 900 kDa band resolving at longer
exposures (ii). Inclusion of DDM and Triton detergent appeared to promote recovery
of monomeric (280 kDa) LRRK2 as the relative band intensity of the 280 kDa species
was greater for both Triton and DDM containing samples relative to detergent free
extracts. In addition, DDM produced a V5 immunoreactive 900 kDa species at short
blot exposures (bands absent in detergent free extracts at these exposure times)
with the detergent possibly promoting the recovery of the higher molecular weight
LRRK2 species. However due to the sample smearing encountered at longer blot
exposures we concluded that the conditions obtaining optimal signal to noise ratios
in LRRK2-V5 over-expressing SHSY5Y cells were those which omitted the use of
detergent.
BN PAGE and Western blot analysis of detergent free, soluble extracts of
human and wild type but not LRRK2 knockout mouse cerebellum revealed a weak
280 kDa 3514-1 immunoreactive band (←) co-migrating with wild type LRRK2-V5 in
SHSY5Y over-expressing cells (Fig 3.14B). Higher molecular weight (900 kDa) 3514-1
immunoreactive species were not detected on this blot, suggesting variability
related to sample processing or the Western blot procedure, as has previously been
described for the 3514-1 antibody (Fig 3.1B,C).
138
BN PAGE electrophoresis of cerebellum, cortex, caudate nucleus and medial
putamen regions of the human brain revealed a 3514-1 immunoreactive band of
approximately 280 kDa (i) in all regions analysed, co-migrating with the 280 kDa
band in wild type mouse cerebellum extracts (Fig 3.15, long exposure blots). At long
139
exposures, a 280 kDa 3514-1 immunoreactive band was also detected in the LRRK2
knockout mouse cerebellum, albeit at lower intensity relative to the wild type
mouse cerebellum extract suggesting the antibody detected a non specific species at
the molecular weight of monomeric LRRK2 under the conditions tested. Lower
molecular weight (55-250 kDa) 3514-1 immunoreactive bands of weaker intensities
(ii) were observed in all human brain regions as well as the wild type and LRRK2
knockout mouse cerebellum. Higher molecular weight, 900 kDa 3514-1
immunoreactive bands were also identified in all human and mouse brain samples
including the mouse LRRK2 knockout cerebellum (iii), potentially corresponding to
non-specific binding of the antibody. Multiple poorly resolved bands in both the
mouse and human brain extracts prevented quantitative comparison of endogenous
LRRK2 expression (Fig 3.15). In addition, the specificity of antibodies made it difficult
to identify the true native LRRK2 species in the samples analysed.
140
3.10 Immunohistochemical analysis of LRRK2 protein in brain
Endogenous LRRK2 protein has also previously been detected by
immunohistochemical analysis of wild type mouse and control human brains. To
determine whether the 3514-1 antibody can detect LRRK2 in the human brain by
immunohistochemistry the antibody was initially tested on wild type and LRRK2
knockout mouse cerebellum brain sections. We focused on the cerebellum region of
the mouse brain as it was easy to identify and based on our previous finding of high
cerebellar LRRK2 protein abundance determined by the combined IP and Western
141
blot analysis (Fig 3.12A). Immunohistochemistry was subsequently used to address
the problem of LRRK2 detection in the human brain rather than to study LRRK2 brain
distribution hence other brain regions were not evaluated.
The mouse cerebellum showed the characteristic pattern of neuronal
haematoxyline staining of intricate patterns of folds and fissures. Paraformaldehyde
fixation and subsequent 3514-1 antibody staining of flash frozen wild type mouse
brain tissue revealed positive DAB immunoreactivity in the Purkinje neurons of the
cerebellum (Fig 3.16A, B). Equivalent regions in the LRRK2 knockout mouse brain did
not show DAB positivity when processed in parallel to wild type mouse brain
sections suggesting DAB staining represented LRRK2. Diluting the concentration of
the 3514-1 antibody (1:50-500) revealed the most intense DAB immunoreactivity for
the 1:50 antibody dilution with lower intensity DAB staining a lower antibody
concentrations (Fig 3.16 C). The pattern of 3514-1 DAB staining in the cerebellar
sections, with purkinje neuron positivity was maintained across all antibody
concentrations (1:50-500).
142
3514-1 staining (1:100 dilution) of flash frozen human cerebellar sections
produced DAB positive staining in the purkinje neurons, comparable to the staining
observed for wild type mouse cerebellum suggesting that LRRK2 protein is present in
this cell population (Fig 3.17). Our results confirm that LRRK2 protein is expressed in
143
the human cerebellum at least, despite our inability to detect LRRK2 by Western blot
or by the immunoprecipitation and Western blot approach.
144
Chapter 4: Results
LRRK2 and mitochondrial function
Mitochondrial abnormalitites that have previously been linked to PD in
genetic and environmental PD models include reduced respiratory chain enzyme
activity [97-100], limited substrate supply to oxidative phosphorylation due to
defects in the glucose transport system [172] and ATPase reversal in the presence of
gross mitochondrial respiratory chain defects [172]. These reported defects have
correlated with reduced activity of the mitochondrial respiratory chain and elevated
ROS generation [107, 172]. Furthermore, mitochondrial localization of PD associated
proteins; PINK1, parkin, DJ1 and α-synuclein have been linked to mitochondrial
functional changes associated with their respective mutant proteins.
Having characterised LRRK2 expression in fibroblast, lymphoblast and
SHSY5Y cell models in Chapter 3, this chapter evaluated mitochondrial function in
wild type and G2019S LRRK2 expressing fibroblasts and SHSY5Y cells by measuring
cellular oxygen consumption, mitochondrial membrane potential, glycolytic
metabolism, basal ATP content and ROS generation. In addition, the underlying
mechanisms of compromised mitochondrial energetics were explored; the
relationship between mitochondrial changes, LRRK2 mitochondrial localization and
kinase activity was evaluated in wild type and G2019S LRRK2-V5 over-expressing
SHSY5Y cells in the presence and absence of LRRK2 kinase inhibition. The potential
correlation between LRRK2 levels and mitochondrial function were assessed in LRRK2
over-expressing cells and the LRRK2 siRNA knockdown SHSY5Y model.
4.1 Mitochondrial function in fibroblast cell lines
4.1.1 The analysis of mitochondrial respiration in control and G2019S fibroblasts
Cellular oxygen consumption was evaluated in intact control and G2019S
fibroblasts by measuring extracellular oxygen concentration using two methods;
145
phosphorescent oxygen probes and polarographically with the Clark type oxygen
electrode.
Phosphorescent measurements were carried out on fibroblasts cultured
under standard conditions (37 ËšC, 5 % CO2) in glucose rich medium with respiration
rates measured over the period of one hour. Phosphorescent analysis of oxygen
consumption in control fibroblasts revealed comparable respiratory rates (0.16-0.18
RFU/min/µg protein) for the six cell lines (Fig 4.1A) while substantial variability in
cellular respiration rates was observed in the 8 G2019S LRRK2 fibroblast lines (0.180.6 RFU/min/µg protein). Comparing the fibroblast cell lines as control versus
G2019S groups revealed a 128 % (p<0.0005) increase in cellular respiration rates for
the G2019S cohort (Fig 4.1B).
Polarographic measurements of oxygen consumption in the presence of
glucose identified a 76 % (p<0.05) increase in respiration for the G2019S fibroblast
group (n=8) relative to control (n=6) cell lines (Fig 4.1C). The increase in respiration
identified in G2019S LRRK2 fibroblasts by the phosphorescent analysis and
polarographic measurements is consistent with an up-regulation of mitochondrial
function in these cell lines relative to control cells.
Polarographic methods were also used to assess maximal respiratory rates
in control and G2019S fibroblasts in response to addition of the chemical uncoupler,
FCCP. FCCP induced a two fold increase in the basal oxygen consumption for control
cells (Fig 4.1C). In G2019S cell lines FCCP addition increased the rate of oxygen
consumption to that seen in FCCP treated control fibroblasts indicating no change in
the maximal mitochondrial capacity. However, ratios of maximal to basal respiration
were reduced compared to control cells suggesting mitochondria may be more
uncoupled in LRRK2 G2019S cells.
146
4.1.2 The analysis of mitochondrial content in control and G2019S LRRK2
fibroblasts
Cellular mitochondrial content can impact on cellular respiratory capacity
therefore
interpretation
of
oxygen
consumption
measurements
required
assessment of mitochondrial abundance. Mitochondrial content was quantified by
Western blot analysis of various mitochondrial proteins in whole cell lysates of
control and G2019S LRRK2 fibroblasts and the proportion of cells stained with
TMRM. Relative to control fibroblasts the levels of inner mitochondrial membrane
147
localized Complex III (Core CIII) and II (SDHA) subunits as well as the mitochondrial
transcription factor (TFAM) were similar in G2019S cells as determined by Core
Complex III, SDHA and TFAM immunoreactivities (Fig 4.2A). Actin immunoreactivity
demonstrated equivalent protein loading. These results are consistent with
comparable electron transport chain subunit densities and abundance of
mitochondrial DNA transcriptional regulators.
Confocal analysis and quantification of the cellular proportion of control and
G2019S LRRK2 fibroblasts stained with TMRM revealed comparable TMRM volumes
per cell (20%) consistent with similar levels of mitochondria in control and G2019S
cell lines (Fig 4.2B). Both analyses suggest LRRK2 G2019S expression in fibroblasts
did not affect mitochondrial levels in fibroblasts.
148
4.1.3 The analysis of mitochondrial membrane potential in control and G2019S
LRRK2 fibroblasts
Mitochondrial membrane potential was assessed in 6 control and 8 G2019S
fibroblasts cultured in glucose rich respiration buffer by TMRM staining and confocal
microscopy. The intensity of TMRM staining appeared reduced in G2019S fibroblasts
relative to control cells (Fig 4.3A). Quantification of TMRM intensity revealed
comparable average values (1150-1200 RFU) for all the cell lines in the control group
(Fig 4.3B). TMRM fluorescence was decreased for all G2019S fibroblast lines by 20 %
(p<0.0001) relative to control cells (Fig 4.3C), consistent with a reduction in the
mitochondrial membrane potential.
149
4.1.4 The analysis of cellular ATP content in control and G2019S LRRK2 fibroblasts
Basal ATP content was assessed in 6 control and 4 G2019S cell lines by
measuring luciferase luminescence in cellular extracts. Considerable variability was
observed in luciferase luminescence for control (3.7x106 -4.8x106 RFU/µg protein)
150
and G2019S mutant fibroblasts (3.6 x106 – 4.4 x106 RFU/µg protein) (Fig 4.4A).
Statistical comparison of control and G2019S fibroblasts identified a 10 % (p<0.05)
decrease in average luminescence for the G2019S cells (Fig 4.4B), consistent with a
reduced basal ATP content. The reduction in ATP content for the four G2019S cell
lines did not correlate with changes in oxygen consumption rates (Fig 4.1A).
4.1.5 The analysis of extracellular acidification rates in control and G2019S LRRK2
fibroblasts
Extracellular acidification rates were assessed in 6 control and 8 G2019S
LRRK2 fibroblasts. Mesurements were carried out on whole cells in glucose rich,
buffer free culturing conditions (37 ËšC) using pH sensitive phosphorescent probes
over the period of one hour. Analysis of pH probe fluorescence revealed
considerable variation in phosphorescent lifetime rate measurements for the
control (1.2 x10-10-2.4 x10-10 [H+]/min/µg protein) and G2019S (0.3 x10-10 -4 x10-10
151
[H+]/min/µg protein) fibroblast groups (Fig 4.5A) suggesting variable rates of
anaerobic glycolysis. Statistical comparison of control versus G2019S cell groups
showed comparable extracellular acidification rates despite the variability observed
among the individual cell lines (Fig 4.5B). Data shown indicates there was no obvious
perturbation in glycolytic rates or correlation with changes in oxygen consumption in
the G2019S fibroblasts.
4.1.6 The analysis of cellular ROS in control and G2019S LRRK2 fibroblasts
To determine whether ROS levels were altered in G2019S expressing
fibroblasts relative to control cells, ROS generation was evaluated by measuring
dihydroethidium oxidation. Dihydroethidium oxidation was monitored in 3 control
and 3 G2019S fibroblast cell lines as the rate of increase in red fluorescence. The
average rate of nuclear accumulation of red DHE fluorescence was 36 % (p<0.005)
152
slower in G2019S cell lines relative to control fibroblasts (Fig 4.6) consistent with
reduced rates of reactive oxygen species generation in the LRRK2 mutant cell lines.
4.1.7 The analysis of mitochondrial proton leak in control and G2019S LRRK2
fibroblasts
To further investigate the possibility of mitochondrial uncoupling,
mitochondrial proton leak was evaluated by measuring oxygen consumption rates
polarographically in the presence of oligomycin for control and G2019S fibroblasts.
Polarographic measurements of oxygen consumption rates in 4 control fibroblasts in
the presence of glucose detected a respiratory rate comparable to the previous
measurement (Fig 4.1A). The addition of oligomycin resulted in ablation of
mitochondrial respiration in control cells (Fig 4.7), an effect that has previously been
associated with complex V inhibition. In contrast, respiration rates detected in
oligomycin treated G2019S fibroblasts were 14 % of untreated rates (p<0.05)
consistent with a mitochondrial proton leak.
153
4.1.8 The analysis of mitochondrial morphology in control and G2019S LRRK2
fibroblasts
Mitochondrial morphology has previously been reported to correlate with
functional changes of the mitochondrial respiratory chain, [119, 121]. Mitochondrial
morphology was assessed in 6 control and 8 G2019S LRRK2 cells by confocal
microscopy focusing on the pattern of TMRM staining. Fibroblasts were imaged in
respiration buffer following incubation with TMRM. In control fibroblast cells TMRM
stained mitochondrial structures consisted of a web of interconnected
mitochondrial processes, dense around the nuclear contour and stretching
throughout the cell body (Fig 4.8). Both long (0.5 µm, ↓) and short (0.025 µm, ↓*)
mitochondrial fragments could be identified representing elongated and fragmented
regions of the mitochondrial network.
154
The G2019S fibroblast cell lines presented with a TMRM staining pattern
resembling that of control cells consistent with a similar architecture of
mitochondrial networks (Fig 4.8). The pattern of TMRM staining was similar for the
individual cells in each control and G2019S fibroblast cell line, consistent with a
uniform mitochondrial distribution and structure.
155
4.2 Mitochondrial dysfunction models
To exclude other potential causes that could account for the observed
G2019S changes, control fibroblast cell lines were used to model; respiratory chain
enzymatic activity defects, limited substrate supply and mitochondrial ATPase
reversal to determine how the bioenergetic changes associated with these models
compare to changes identified in G2019S expressing fibroblasts.
4.2.1 The use of control fibroblasts treated with rotenone or antimycin A as a
model of respiratory chain dysfunction
Respiratory chain enzymatic activity defects, such as those reported in
idiopathic PD patients, α-synuclein over-expressing SHSY5Y cells and PINK1 mutant
fibroblasts from PD patients were modelled by treating control fibroblasts with
inhibitors of Complex I (rotenone 1 µM) and complex III (antimycin A, 1 µM). Oxygen
consumption rates of control fibroblasts, treated or untreated with ETC inhibitors
were assessed polarographically in the presence of glucose. A 70 % reduction in
phosphorescent lifetime rate values was detected for control fibroblasts treated
with 1 µM rotenone relative to untreated cells, reflecting reduced respiratory chain
activity (Fig 4.9). Oxygen consumption was not completely inhibited, as shown by
the 30 % basal respiration still taking place in fibroblasts treated with rotenone,
potentially reflecting compensatory effects of increased substrate supply through
complex II of the respiratory chain. The rate of change of phosphorescent lifetime
was reduced to 30 % of the basal rate for untreated fibroblasts with the addition of
antimycin A, comparable to the response of fibroblasts to rotenone (Fig 4.9).
156
4.2.2 The use of control fibroblasts treated with rotenone and oligomycin to
evaluate the influence of ETC inhibition on the mitochondrial ATPase
Mitochondrial ATPase reversal has previously been associated with PINK1
and parkin linked PD as well as electron transport chain response to enzymatic
inhibition. When complex I is inhibited the mitochondrial membrane potential is
maintained by the reversal of the ATPase [310]. Control fibroblasts were treated
with rotenone (1 µM) and the mitochondrial proton gradient monitored in
respiration buffer by confocal microscopy of TMRM stained cells. Quantification of
TMRM intensity of rotenone treated control fibroblast revealed fluorescent intensity
values comparable to those of untreated cells 3 minutes after administration of the
drug (Fig 4.10A). Addition of oligomycin (4 mg/ml) resulted in a time dependent
increase in TMRM fluorescence in untreated control fibroblasts consistent with an
increase in the mitochondrial membrane potential as a result of ATPase inhibition
and accumulation of protons in the mitochondrial matrix. In contrast, oligomycin
157
addition to rotenone treated fibroblasts resulted in an immediate 30 % decrease in
TMRM fluorescent intensity consistent with the dissipation of the mitochondrial
proton gradient.
TMRM fluorescent intensity in G2019S fibroblast cell lines was stable for 6
minutes following addition of oligomycin (4 mg/ml), comparable to oligomycin
effects on control cells. Hence there is no evidence to suggest the mitochondrial
membrane potential is being maintained in these cells by reversal of the ATPase (Fig
4.10B).
158
4.2.3 The use of control fibroblasts cultured in galactose rich conditions as a model
of limited cellular substrate supply
Oxygen consumption rates were assessed in control fibroblasts by
phosphorescent analysis under glucose or galactose culture conditions (37 ËšC, 5 %
CO2) over the course of one hour. As the rate limiting enzyme in galactose
metabolism has slow first order rate kinetics, glycolytic flux is restricted.
Phosphorescent probe lifetime rate measurements of control fibroblasts cultured in
159
galactose medium were 1.5 fold (p<0.0005) greater relative to rates recorded for
control fibroblasts cultured in glucose consistent with an increase in cellular oxygen
consumption when galactose was supplied as a substrate source (Fig 4.11A).
Extracellular acidification rates were assessed for control fibroblasts in
glucose or galactose rich, buffer free culture conditions (37 ËšC, 5 % CO2) using pH
sensitive phosphorescent probes over the period of one hour. Control fibroblasts
cultured in glucose showed a linear rate of increase in extracellular proton
concentrations, consistent with a steady rate of anaerobic glycolysis and lactate
generation (Fig 4.11B). Monitoring extracellular acidification rates in control
fibroblasts cultured in galactose, revealed 10 fold slower (p <0.0005) extracellular
acidification rates relative to fibroblasts cultured on glucose consistent with a
decreased rate of glucose metabolism.
The mitochondrial membrane potential was assessed in control fibroblasts
cultured in glucose or galactose rich respiration buffer by confocal microscopy of
TMRM stained cells. Quantification of TMRM fluorescent intensity identified a 10 %
(p<0.0001) increase in TMRM fluorescence of fibroblasts cultured in galactose
medium when compared to cells cultured under glucose conditions (Fig 4.11C)
consistent with an increase in the mitochondrial membrane potential.
Basal ATP levels were evaluated in control fibroblasts cultured in glucose or
galactose rich conditions by quantification of luciferase luminescence in lysed cells.
Luminescence was comparable in cells cultured on glucose or galactose (Fig 4.11D),
suggesting basal ATP levels were unaffected by restriction in glycolytic substrate
supply.
160
4.3 The analysis of respiratory chain function in the SHSY5Y LRRK2-V5 overexpressing model
To confirm the defect in fibroblasts was replicated in SHSY5Y cells overexpressing G2019S LRRK2 mitochondrial function was evaluated in the SHSY5Y cell
model. Control, wild type and G2019S LRRK2-V5 over expressing SHSY5Y cells were
assessed for oxygen consumption rates by polarographic measurements,
mitochondrial membrane potential by TMRM staining and confocal microscopy and
161
reactive oxygen species by spectrophotometric quantification of cellular aconitase
activity.
Polarographic measurements of oxygen consumption by the Clark type
oxygen electrode revealed basal respiration in control SHSY5Y cells (4 nmol
O/min/µg protein) comparable to rates determined for wild type LRRK2-V5 overexpressing SHSY5Y cells (Fig 4.12A). In contrast SHSY5Y cells over-expressing G2019S
LRRK2-V5 had an 80 % (p<0.05) increase in the rate of oxygen consumption relative
to control and wild type LRRK2-V5 over-expressing SHSY5Y cells (Fig 4.12A).
Quantification of TMRM intensity determined comparable intensity measurements
for control and wild-type LRRK2-V5 over-expressing SHSY5Y cells while a 20 %
(p<0.0001) reduction in TMRM intensity was identified for G2019S LRRK2-V5 overexpressing SHSY5Y cells (Fig 4.12B), consistent with a decrease in the mitochondrial
membrane potential. Comparable aconitase activities were determined for control,
wild type and G2019S LRRK2-V5 over-expressing SHSY5Y cell lines consistent with
the absence of oxidative stress (Fig 4.12C).
162
4.4 LRRK2 kinase activity
4.4.1 The analysis of LRRK2 kinase inhibition in the SHSY5Y LRRK2-V5 overexpressing model
LRRK2 kinase activity was previously shown to be greater for G2019S LRRK2V5 immunoprecipitated from SHSY5Y over-expressing SHSY5Y cells relative to the
wild type protein (Fig 3.8C). To determine the influence of LRRK2 kinase inhibition
on Serine 935 phosphorylation and mitochondrial function in SHSY5Y cells, LRRK2
163
over-expressing SHSY5Y cells were subject to treatment with LRRK2 specific kinase
inhibitors. LRRK2 was immunoprecipitated from the cells and assessed by Western
blot analysis using phospho Serine 935 specific LRRK2 antibodies. Mitochondrial
function was assessed polarographically and by TMRM staining following LRRK2
kinase inhibition.
4.4.1.1 LRRK2-IN1
LRRK2 was immunoprecipitated from wild type LRRK2-V5 over-expressing
SHSY5Y cells treated with increasing concentrations of LRRK2 IN1 (0-2 µM). Western
blot analysis of immunoprecipitates revealed a dose dependent reduction of
phospho Serine 935 immunoreactivity relative to the 3514-1 immunoreactive band
corresponding to the full length immunoprecipitated LRRK2-V5 (Fig 4.13A).
Quantification of the phospho Serine 935 band intensity relative to 3514-1
determined a 90 % reduction in the signal intensity for wild-type LRRK2-V5 SHSY5Y
cells treated with 1µM LRRK2-IN1 when compared to untreated cells, consistent
with a marked inhibition of LRRK2 kinase activity [215]. The effective LRRK2-IN1
concentration (1 µM) was within the range used to dephosphorylate Serine 935 in
HEK293 and lymphoblast cell culture models [215]. Time course analysis (0-90 min)
of wild type LRRK2-V5 over-expressing SHSY5Y cells with 1 µM LRRK2-IN1 treatment
revealed a time dependent reduction in phospho Serine 935 immunoreactivity with
a 90 % decrease in phospho signal relative to 3514-1 band intensity detected
following 90 minutes of treatment when compared to untreated cells (Fig 4.13B), in
line with previous reports.
4.4.1.2 CZC25146
LRRK2 was immunoprecipitated from wild type LRRK2-V5 over-expressing
SHSY5Y cells treated with increasing concentrations of CZC25146 (0-1 µM). Western
blot analysis of immunoprecipitates revealed a dose dependent reduction of
164
phospho Serine 935 immunoreactivity relative to the 3514-1 immunoreactive band
corresponding to the full length immunoprecipitated LRRK2-V5 (Fig 4.13C).
Quantification of the phospho Serine 935 band intensity relative to 3514-1
determined a 90 % reduction in the signal intensity for wild-type LRRK2-V5 SHSY5Y
cells treated with 0.3 µM CZC25146 when compared to untreated cells, consistent
with a marked inhibition of LRRK2 kinase activity. The effective CZC25146
concentration (0.3 µM) was within the range used to dephosphorylate Serine 935 in
HEK293 and lymphoblast cell culture models [216]. In addition, the 0.3 µM
CZC25146 dose resulted in comparable levels of phospho Serine 935
dephosphorylation as the 1 µM dose of LRRK2-IN1 suggesting CZC25146 was more
potent than LRRK2-IN1 in our SHSY5Y cell culture model.
4.4.1.3 TAE-648
LRRK2 was immunoprecipitated from wild type LRRK2-V5 over-expressing
SHSY5Y cells treated with increasing concentrations of TAE-684 (0-1 µM). Western
blot analysis of immunoprecipitated LRRK2 revealed a dose dependent reduction of
phospho Serine 935 immunoreactivity relative to the 3514-1 immunoreactive band
corresponding to the full length immunoprecipitated LRRK2-V5 (Fig 4.13D).
Quantification of the phospho Serine 935 band intensity relative to 3514-1
determined a 90 % reduction in the signal intensity for wild-type LRRK2-V5 SHSY5Y
cells treated with 0.3 µM TAE-684 when compared to untreated cells, consistent
with a marked inhibition of LRRK2 kinase activity. The effective TAE-684
concentration (0.3 µM) was within the range used to dephosphorylate Serine 935 in
HEK293 and lymphoblast cell culture models [214]. In addition, the 0.3 µM TAE-684
dose resulted in comparable levels of phospho Serine 935 dephosphorylation,
relative to total LRRK2 as the 1 µM dose of LRRK2-IN1 suggesting TAE-684 was a
more potent LRRK2 kinase inhibitor in our cell culture model.
165
166
4.4.2 The use of LRRK2 kinase inhibitors to evaluate effects of LRRK2 kinase
inhibition on mitochondrial function in the SHSY5Y LRRK2-V5 over-expressing
model
4.4.2.1 Evaluation of the effects of LRRK2 kinase inhibition on mitochondrial
respiration
Polarographic measurements of cellular oxygen consumption in the
presence of glucose carried out on control, wild type and G2019S LRRK2-V5 cells or
cells treated with LRRK2-IN1 (1 µM, 90 min) revealed comparable respiration rates
for control and wild type LRRK2-V5 over-expressing SHSY5Y cells in the presence and
absence of LRRK2-IN1 (Fig 4.14A) suggesting LRRK2 kinase inhibition did not
influence mitochondrial respiration. In contrast, a 25 % (p<0.05) reduction in oxygen
consumption rate was reported for G2019 LRRK2-V5 over-expressing SHSY5Y cells
following LRRK2 kinase inhibition with LRRK2-IN1. The inhibited rates were
comparable to untreated control and wild type LRRK2-V5 over-expressing SHSY5Y
cells (Fig 4.14A). LRRK2 Serine 935 phosphorylation status was evaluated following
the respiratory measurements. Phospho Serine 935 immunoreactivity relative to
3514-1 band intensity, corresponding to total LRRK2-V5 levels, was reduced by >80
% for LRRK2-IN1 treated wild type and G2019S LRRK2-V5 over-expressing SHSY5Y
cells relative to untreated cells (Fig 4.14B), consistent with a marked inhibition of
LRRK2 kinase activity.
4.4.2.2 Evaluation of the effects of LRRK2 kinase inhibition on the mitochondrial
membrane potential
Control, wild type and G2019S LRRK2-V5 cells were treated with LRRK2-IN1
(1 µM, 90 min), CZC25146 (0.3 µM) or TAE-684 (0.3 µM) following which the
mitochondrial membrane potential was assessed in untreated or inhibitor treated
cells by confocal microscopy of TMRM stained cells in glucose rich respiration buffer.
167
LRRK2-IN1 treatment of wild type and G2019S LRRK2-V5 over-expressing SHSY5Y
cells resulted in 10 % and 40 % increases (p<0.0001) in TMRM fluorescence relative
to untreated cells (Fig 4.14C) suggesting LRRK2 kinase inhibition increased the
mitochondrial membrane potential.
CZC25146 (0.3 µM) treatment of control and wild type LRRK2-V5 overexpressing SHSY5Y cells resulted in a 10 % (p<0.0001) decrease in the TMRM
fluorescence relative to untreated cells consistent with a reduction in the
mitochondrial membrane potential. CZC25146 effects on the TMRM intensity for
wild type LRRK2-V5 over-expressing SHSY5Y cells contrasted the 10 % TMRM
intensity increase observed with LRRK2-IN1 treatment (Fig 4.14D), potentially due to
the different off target effects described for the two compounds. In contrast,
CZC25146 treatment of G2019S LRRK2-V5 over-expressing cells resulted in a 10 %
(p<0.0001) increase in the intensity of TMRM staining consistent with an increase in
the mitochondrial membrane potential. Although opposite effects were observed
for control and wild type LRRK2-V5 over-expressing SHSY5Y cells in response to
LRRK2-IN1 and CZC25146 treatment, the increase in TMRM intensity for the G2019S
LRRK2-V5 expressing cells following kinase inhibition suggest LRRK2 kinase inhibition
potentially restores the mitochondrial abnormality associated with G2019S LRRK2V5 protein.
TAE-684 (0.3 µM) treatment of control, wild type and G2019S LRRK2-V5
over-expressing SHSY5Y cells resulted in a 60 % (p<0.0001) decrease in TMRM
fluorescence relative to untreated cells suggesting TAE-684 could potentially inhibit
mitochondrial function, possibly due to the broader selectivity profile of the drug
(Fig 4.14D).
168
4.5 LRRK2 mitochondrial localization
4.5.1 Mitochondrial affinity purification
Cellular fractionation experiments carried out in chapter 1 identified 10 % of
the total LRRK2 cellular pool in the mitochondrial enriched fractions (MEF) for wild
type and G2019S LRRK2-V5 over-expressing SHSY5Y cells as well as endogenous
LRRK2 in lymphoblasts.
169
To determine whether LRRK2 localizes to the mitochondria and quantify the
mitochondrial LRRK2 pool, mitochondrial affinity purification was carried out on wild
type, G2019S LRRK2-V5 over-expressing cells and lymphoblasts. Affinity purified
mitochondria were isolated from the post nuclear supernatant (PNS) of cells and
assessed by Western blot analysis. A fraction (1/10th) of the PNS sample was
retained for analysis of LRRK2 and mitochondrial input with all Western blot
quantifications corrected for sample loading: PNS (1/10) and purified mitochondria
(9/10) unless stated otherwise.
Affinity purified mitochondria were identified by Western blot analysis of
the 39 kDa immunoreactive band for Core Complex III marker of the inner
mitochondrial matrix, quantified as 11 fold greater signal intensity relative to the
Core Complex III in the PNS input fraction (Fig 4.15A). Correcting the Core Complex
III immunoreactive signal for sample loading calculated >90 % recovery of
mitochondria from the PNS input. Western blot immunoreactivity for cytochrome C
(12 kDa), a marker of the inter membrane space, was 2 fold lower relative to the
signal detected in the PNS input suggesting a portion of cytochrome C was lost
during the affinity purification procedure. Western blot analysis detected calreticulin
and GCase immunoreactivities in the PNS fraction of SHSY5Y cells but were absent in
extracts of affinity purified mitochondria suggesting they were generally free from
lysosomal and ER contaminants. V5 immunoreactivity was detected in the PNS and
mitochondrial affinity purified fractions. Quantification of the mitochondrial V5 band
intensity relative to PNS and corrected for sample loading determined 1-2 % (n=3) of
the post-nuclear V5 immunoreactivity in the affinity purified fraction consistent with
a mitochondrial localization of LRRK2-V5.
170
4.5.1.1 Mitochondrial affinity purification versus differential centrifugation
Direct comparison of LRRK2 recovery from mitochondria isolated by
differential centrifugation or affinity purification identified a 39 kDa immunoreactive
band of 10 fold greater signal intensity relative to the PNS input lysates of wild type
LRRK2-V5 over-expressing SHSY5Y cells, suggesting comparable mitochondrial
recoveries by the two purification methods (Fig 4.15B). While mitochondrial
enriched fractions isolated by differential centrifugation were immunoreactive for
GCase, affinity purified preparations did not reveal clear GCase immunoreactivity in
the same exposure time consistent with low lysosomal abundance. V5
immunoreactivity in mitochondria isolated by differential centrifugation was of 2
fold lower intensity relative to V5 in PNS input consistent with 7 % (n=1) of the PNS
LRRK2 pool present in this mitochondrial isolation when corrected for sample
loading (1/15th of the PNS was retained for Western blot analysis in this experiment).
A V5 immunoreactive signal of 6 fold lower intensity was detected in affinity purified
preparations consistent with 1% of the PNS LRRK2 pool in this mitochondrial
fraction, consistent with previous observations (Fig 4.15A). Our data suggests the
LRRK2 signal in the mitochondrial enriched fraction of crude preparations is
potentially overestimated.
4.5.1.2 Mitochondrial affinity purification from G2019S LRRK2-V5 over-expressing
SHSY5Y cells
The 280 kDa immunoreactive V5 band was detected in mitochondrial affinity
purifications from wild type and G2019S LRRK2-V5 over-expressing SHSY5Y cells.
Quantification of wild type or G2019S V5 immunoreactivity in the affinity purified
mitochondrial fraction relative to PNS and corrected for sample loading determined
1.5 and 1.6 % (n=2) of the V5 signal in the purified mitochondrial compartment
consistent with comparable levels of LRRK2 mitochondria isolated from wild type
171
and G2019S LRRK2-V5 over-expressing SHSY5Y cells (Fig 4.15C). Mitochondrial
affinity preparations from SHSY5Y cells over-expressing wild type or G2019S LRRK2V5 were free of lysosomal contaminants as determined by the absence of a clear
GCase immunoreactive signal as detected in the PNS input.
4.5.1.3 Mitochondrial affinity purification, LRRK2 kinase inhibition
To
determine
whether
LRRK2
phosphorylation
status
affects
its
mitochondrial localization, SHSY5Y wild type over-expressing LRRK2-V5 cells were
treated with LRRK2 IN1 (1 µM, 90 min) and affinity purified mitochondrial fractions
assessed for relative LRRK2-V5 abundance by Western blot. Inhibition of LRRK2
kinase activity and Serine 935 de-phosphorylation did not affect the efficiency of
mitochondrial recovery as determined by comparable Core Complex III
immunoreactivities for LRRK2-IN1 treated and untreated wild type LRRK2-V5 overexpressing SHSY5Y cells relative to the PNS when corrected for sample loading (Fig
4.15D). The 280 kDa immunoreactive V5 band was detected in mitochondrial affinity
purifications from LRRK2-IN1 treated and untreated wild type LRRK2-V5 cells.
Quantification of the V5 band intensity relative to PNS input and sample loading
determined 1.6 and 1.7 % (n=2) of V5 immunoreactivity in untreated and LRRK2-IN1
treated SHSY5Y cells consistent with comparable LRRK2 mitochondrial abundance in
the presence and absence of LRRK2 kinase inhibition and Serine 935
dephosphorylation.
4.5.1.4 Mitochondrial affinity purification from lymphoblasts
Affinity purified mitochondria were isolated from lymphoblasts to
determine the proportion of endogenous protein in the mitochondrial compartment
of this cell type. Successful purification of mitochondria was confirmed by the
presence of a 39 kDa immunoreactive Core Complex III band in the affinity purified
mitochondrial preparation (Fig 4.15E). Correcting the Core Complex III
172
immunoreactivity for sample loading relative to Core Complex III band intensity in
the PNS input determined >90 % mitochondrial recovery by this purification method,
comparable to purification recovery obtained with SHSY5Y cells (Fig 4.15A, C). Weak
3514-1 280 kDa immunoreactivity was detected by Western blot analysis of
lymphoblast PNS input, however, 3514-1 immunoreactivity could not be clearly
identified in affinity purified mitochondrial preparations (Fig 4.15E). In an attempt
to increase the sensitivity of endogenous LRRK2 detection, LRRK2 was
immunoprecipitated from the PNS input and affinity purified mitochondria. A clear
280 kDa 3514-1 immunoreactive band was identified in the lymphoblast PNS input
(Fig 4.15E). LRRK2 immunoprecipitation from mitochondrial isolates containing 10
times the volume of the PNS input detected a faint 3514-1 immunoreactive band comigrating with the band identified in the PNS input consistent with low levels
mitochondrial LRRK2 in lymphoblasts. The background noise did not allow for
quantification of the relative LRRK2 mitochondrial abundance in lymphoblasts.
173
174
4.5.2.1 Sub-fractionation of mitochondrial enriched fractions from SHSY5Y LRRK2V5 over-expressing SHSY5Y cells
To determine the intra mitochondrial distribution of LRRK2, mitochondrial
fractionation was carried out in wild type LRRK2-V5 over-expressing SHSY5Y cells.
Mitochondrial sub-fractionation was initially carried out on mitochondria isolated by
differential centrifugation before extending the studies to affinity purified
mitochondria. The outer mitochondrial membrane of mitochondria was selectively
removed by digitonin treatment. Following treatment, mitochondrial enriched
fractions were separated into soluble and pellet components by centrifugation and
the corresponding extracts analysed by Western blotting. Western blot analysis of
untreated MEFs identified porin (outer mitochondrial membrane marker) and Core
Complex III (inner mitochondrial membrane marker) immunoreactivities in the
pelleted mitochondrial compartment absent in the corresponding soluble fraction
(Fig 4.16A). V5 immunoreactivity was also detected in the untreated pellet of MEFs
in line with our previous findings (Fig 4.15B).
MEFs isolated from wild type LRRK2-V5 over-expressing SHSY5Y cells treated
with 2 mg/ml digitonin revealed 20 % of the total porin immunoreactivity in the
soluble fraction with 80 % retained in the mitochondrial pellet. At 3 mg/ml porin
immunoreactivity in the soluble fraction was comparable to the band intensity
detected in the mitochondrial pellet consistent with the removal of 50 % of the
outer mitochondrial membrane. Core Complex III immunoreactivity was absent in
the soluble fraction following 2 mg/ml digitonin treatment but 50 % of the total Core
Comlex III immunoreactivity was found in the soluble compartment at 3 mg/ml
digitonin, consistent with the increased solubilisation of the inner mitochondrial
membrane at this digitonin concentration. At 3 mg/ml, when the inner
mitochondrial membrane of MEF preparations was detected in the soluble fraction
175
60 % of total porin immunoreactivity was retained in the MEF pellet suggesting a
portion of the outer membrane was still associated with the inner membrane even
though the integrity of the inner membrane was compromised.
Western blot analysis of untreated MEFs showed that V5 immunoreactivity
was absent in the soluble fraction while MEFs treated with 1 mg/ml digitonin had 50
% of the total V5 immunoreactivity in the soluble fraction consistent with
solubilisation of half of the mitochondrial LRRK2 pool. Mitochondria treated with 2
mg/ml digitonin revealed 60 % and 40 % V5 immunoreactivities in the soluble and
mitochondrial pellet fractions while at 3 mg/ml digitonin the V5 immunoreactive
signal was equally distributed between the two fractions. Although we have
previously demonstrated 90 % of the LRRK2 immunoreactivity in mitochondria
isolated by differential centrifugation potentially represents signal contamination
from other cellular fractions, data shown suggests LRRK2 response to digitonin
mimics that of porin consistent with LRRK2 localization to the outer mitochondrial
membrane of MEFs.
4.5.2.2 Sub-fractionation of affinity purified mitochondria from SHSY5Y LRRK2-V5
over-expressing SHSY5Y cells
To determine LRRK2 localization in affinity purified mitochondria from wild
type LRRK2-V5 over-expressing SHSY5Y cells, affinity purified mitochondrial
preparations were subject to digitonin treatment. Western blot analysis of
untreated affinity purified mitochondria identified 60 kDa porin and 39 kDa Core
Complex III and V5 immunoreactivities in the mitochondrial pellet, absent in the
soluble fraction, consistent with intact mitochondrial inner and outer membranes
(Fig 4.16B). Digitonin at 1 mg/ml resulted in 10 % of total porin immunoreactivity in
the soluble fraction while 2 mg/ml digitonin the soluble pool was estimated to
contain approximately 20 % porin immunoreactivity. Higher concentrations, 3 and 4
176
mg/ml produced 40 % porin immunoreactivity in the soluble fraction with 60 %
immunoreactivity retained in the mitochondrial pellet consistent with a dose
dependent solubilisation of the outer mitochondrial membrane.
Western blot analysis of affinity purified mitochondria treated with digitonin
at 1 mg/ml detected 5 % of total Core Complex III immunoreactivity in the soluble
fraction (Fig 4.16B). Digitonin at 2 mg/ml resulted in 20 % of the Core Complex III
signal intensity present in the soluble fraction, while 3 and 4 mg/ml increased the
soluble Core Complex III signal intensity to 50 % consistent with a dose dependent
solubilization of the inner mitochondrial membrane. Similar to the data obtained for
sub-cellular fractionation of mitochondria isolated by differential centrifugation, the
digitonin titration did not appear to selectively target the outer mitochondrial
membrane in highly purified mitochondrial preparations as the inner membrane was
solubilised while a large portion of the outer membrane appeared intact. Despite
the unpredictable inner and outer mitochondrial membrane marker behaviour,
LRRK2 V5 immunoreactivity signal reflected the pattern of porin immunoreactive
distribution with 10 %, 10 % and 20 % of the total V5 signal intensity identified in the
soluble fraction for 1, 2 and 3 mg/ml digitonin, consistent with an outer
mitochondrial membrane localisation in affinity purified mitochondrial preparations
from wild type LRRK2-V5 over-expressing SHSY5Y cells.
Attempts to sub-fractionate mitochondrial preparations with digitonin did
not give the predicted data even after numerous attempts hence an alternative
approach was tested. Affinity purified mitochondria were digested with proteinase k
(5 and 10 µg/ml) for increasing incubation periods (0-20 min) following which
mitochondria were pelleted by centrifugation and assessed by Western blot analysis.
At 5 µg/ml proteinase, 39 kDa immunoreactivity for Core Complex III was 2 fold
lower relative to untreated mitochondria for all the digestion times, consistent with
177
a partially digested inner mitochondrial membrane (Fig 4.16C). Increasing proteinase
K to 10 µg/ml resulted in a 4 fold reduction of Core Complex III immunoreactivity for
all time points analysed consistent with a correlation between enzyme
concentration and the rate of mitochondrial inner membrane digestion by
proteinase K. Porin immunoreactivity was present in untreated mitochondria but
was absent in the presence of proteinase K for all concentrations and time points
tested with a faint immunoreactive band detected following 20 minute treatment
with 5 µg/ml proteinase K. The absence of a porin immunoreactive signal suggested
the outer mitochondrial membrane was completely digested even after 5 minutes of
proteinase K incubation at 5 µg/ml. An additional 15 kDa immunoreactive porin
species was identified in affinity purified mitochondria treated with 5 and 10 µg/ml
proteinase K for all time points (0-20 min) analysed. The 15 kDa porin
immunoreactivity in the presence of the proteinase K was 3 fold greater relative to
the band intensity in untreated mitochondrial preparations, suggesting the band
could represent a porin degradation product.
4.5.3.1 Sub-fractionation of mitochondria to determine LRRK2 membrane
association
Existing data suggests LRRK2 is more abundant in a membrane associated
form, preferentially binding to cholesterol rich membranes such as those present in
lipid rafts. To determine whether mitochondrial LRRK2 in wild type LRRK2 overexpressing SHSY5Y cells is associated with mitochondrial membranes mitochondria
were fractionated in hypotonic and sodium carbonate buffers. Initially,
mitochondrial sub-fractionation was carried out on MEFs isolated by differential
centrifugation. Following treatment, MEFs were separated into soluble and pellet
components by centrifugation and the corresponding extracts analysed by Western
blotting. Western blot analysis of mitochondria isolated by differential
178
centrifugation identified an immunoreactive signal for the inner membrane marker,
SDHA, in the pelleted MEF while cytochtome C immunoreactivity, representing a
soluble protein in the intermembrane space, was detected in both soluble fraction
(20 %) and pellet (80 %) MEFs (Fig 4.16D). The presence of cytochrome C in the
soluble compartment prior to fractionation suggested the integrity of mitochondria
isolated by differential centrifugation may be partially compromised. LRRK2 was
detected in the mitochondrial pellet as determined by 280 kDa V5 immunoreactivity.
MEFs subject to treatment in hypotonic buffer resulted in a 3 fold increase in
cytochrome C immunoreactivity in the soluble fraction relative to untreated MEFs
consistent with cytochrome C release from the mitochondrial inter membrane space
(Fig 4.16D). The distribution and relative intensity of the SDHA immunoreactivity in
MEFs after the hypotonic treatment resembled that of untreated MEFs suggesting
SDHA remained associated with the mitochondria and the inner mitochondrial
membrane retained intact. V5 immunoreactivity following hypotonic treatment of
MEFs was comparable to that of untreated cells and SHDA suggesting LRRK2 was not
readily released from the mitochondria in hypotonic conditions.
MEFs subject to sodium carbonate treatment resulted in a 4 fold decrease in
the pellet SDHA immunoreactivity with the appearance of an immunoreactive band
in the soluble fraction consistent with the release of the integral mitochondrial
membrane protein in alkaline conditions. LRRK2 immunoreactive distribution in the
presence of sodium carbonate resembled that of SDHA with approximately 30 % of
the V5 signal identified in the soluble fraction characteristic of a membrane
associated protein.
Mitochondrial membrane sub-fractionation of affinity purified mitochondria
was also investigated. V5, SDHA and cytochrome C immunoreactive bands were
identified in the pellet of the untreated affinity purified mitochondria, absent in the
179
soluble fraction consistent with a better preserved mitochondrial integrity relative
to the MEF isolated by differential centrifugation (Fig 4.16E). Cytochrome C was
readily released from the mitochondria under hypotonic conditions as determined
by the appearance of a cytochrome C immunoreactive signal in the soluble fraction
with a corresponding 60 % decrease in the cytochrome C band intensity of the pellet
relative to untreated mitochondria. Mitochondrial SDHA and V5 immunoreactivities
resembled those of untreated mitochondria in hypotonic conditions as shown by the
presence of immunoreactive bands in the pellet extracts only, suggesting both
proteins were still associated with the mitochondria. A weak SDHA and V5
immunoreactive band was detected following sodium carbonate treatment of
mitochondria in the soluble extracts suggesting SDHA and LRRK2 were partially
solubilised from the affinity purified mitochondria in alkaline conditions consistent
with the characteristics of membrane associated proteins.
A large portion of the SDHA and V5 immunoreactive signals were retained in
the pellet fraction. In addition cytochrome C was not fully solubilized with hypotonic
treatment for both affinity purified mitochondria and MEF preparations (Fig 4.16D,
E) as determined by the immunoreactive band in the pelleted fraction suggesting
some mitochondria were still intact. The incomplete solubilisation of membrane
proteins under alkaline conditions and intact mitochondrial membranes made it
difficult to definitively map LRRK2 to mitochondrial membranes.
180
181
Chapter 5: Results
LRRK2 regulation of mitochondrial uncoupling: mechanisms
Mitochondrial permeability to protons can be influenced by the opening of
the mitochondrial transition pore coupled with the release of free radicals [311] or
through reduced integrity of the mitochondrial membrane [312]. Free radical
involvement, mitochondrial morphology and LRRK2 mitochondrial association in
G2019S cells were assessed in Chapter 4. Mitochondrial uncoupling can also occur
through the action of mitochondrial uncoupling proteins [313] which can be
regulated transcriptionally and in the mitochondria by post translational
modification. The PD associated DJ1 protein has been implicated in UCP
transcriptional regulation [168], a pathway which could potentially involve LRRK2. In
addition, mitochondrial bioenergetics have previously been shown to be influenced
by the DLP1 protein, a putative LRRK2 interacting partner [280], suggesting LRRK2
mitochondrial DLP1 association could contribute to the regulation of mitochondrial
uncoupling reported in fibroblasts and SHSY5Y cells. This chapter explores the
potential involvement of UCPs, DJ1 and DLP1 in the regulation of mitochondrial
bioenergetics by LRRK2 in fibroblasts and SHSY5Y cells.
5.1.1 The analysis of SHSY5Y cells over-expressing murine GFP-UCP2
To determine how increased UCP2 expression influences mitochondrial
bioenergetics in SHSY5Y cells, a GFP tagged murine UCP2 construct was transiently
transfected into the cells. Confocal microscopy detected GFP fluorescence in
transfected cells (Fig 5.1A) consistent with GFP-UCP2 protein expression. Co-staining
the cells with the mitochondrial dye, TMRM, confirmed GFP expression in all cells
and the superimposed GFP/TMRM fluorescence showed complete co-localization of
the two fluorophores consistent with mitochondrial localization of the GFP tagged
UCP2 molecule. The pattern of TMRM fluorescence resembled that of untransfected
182
SHSY5Y cells (Fig 5.1B) with no obvious change in mitochondrial morphology
associated with GFP-UCP2 over-expression. Quantification of the TMRM intensity
relative to control cells identified a 20 % (p<0.0001) reduction in GFP-UCP2
expressing cells (Fig 5.1C).
183
5.1.2.1 The analysis of relative UCP mRNA expression in fibroblasts and SHSY5Y
cells
Having identified changes in the mitochondrial membrane potential in
response to UCP2 over-expression, the G2019S expressing fibroblast and SHSY5Y cell
models were screened to quantify relative levels of UCP mRNA and protein. mRNA
was extracted from control SHSY5Y cells and fibroblasts cultured under normal
growth conditions, reverse transcribed to cDNA and UCP expression quantified by
real time PCR analysis as the cycle threshold (CT) value relative to GAPDH CT. Sizes
of all UCP PCR products were confirmed by agarose gel electrophoresis (Appendix
Fig A2). In control fibroblasts UCP1 mRNA was not readily detected while UCP2 and
3 mRNA expression was approximately 5 and 10 fold lower relative to UCP4 and
UCP5 mRNA levels (Fig 5.2A) suggesting UCP5 is the abundant mRNA isoform. In
contrast, UCP2 was the most abundant mRNA isoform in control SHSY5Y cells (Fig
5.2B). UCP3 levels were approximately 6 fold lower when compared to UCP2
expression (Fig 5.2B). UCP4 mRNA abundance was 10 and 3 fold lower relative to
UCP2 and 3. Expression of UCP5 mRNA was 5 and 1.5 fold lower relative to UCP2
and 3 and 2 fold higher relative to UCP4. Our data confirmed UCP2, 3, 4 and 5
expression in both fibroblasts and SHSY5Y cells and showed a cell specific pattern of
UCP mRNA isoform distribution.
184
5.1.2.2 The analysis of relative UCP mRNA expression in G2019S fibroblasts and
SHSY5Y cells
To determine whether UCP2, 3, 4, 5 mRNA expression was altered in cells
expressing G2019S LRRK2, mRNA levels were quantified and compared between
control and G2019S fibroblasts and control, wild type and G2019S LRRK2-V5
expressing SHSY5Y cells. Data was expressed as percentage CT value relative to
control cells. UCP mRNA abundance was higher in the G2019S fibroblasts relative to
control cells (Fig 5.3A) with 300 %, 200 %, 20 % and 15 % increases in the levels of
isoforms 2, 3, 4 and 5 respectively. However, only the UCP2 mRNA increase was
statistically significant. SHSY5Y cells over-expressing wild type LRRK2-V5 had
comparable levels of UCP mRNA to control SHSY5Y cells for all UCP isoforms
analysed (Fig 5.3B), suggesting wild type LRRK2-V5 over-expression did not influence
UCP mRNA levels. However, G2019S LRRK2-V5 over-expressing SHSY5Y cells showed
a 300 % increase in the mRNA levels of the UCP4 isoform with transcript levels of
UCPs 2, 3, 5 expressed at similar abundances relative to control and wild type LRRK2
over-expressing cells (Fig 5.3B) consistent with increased UCP4 mRNA expression.
185
5.1.3 Western blot analysis of UCP2 protein in fibroblasts and SHSY5Y cells
To determine whether increased UCP mRNA expression correlated with UCP
protein, Western blot analysis was carried out on whole cell lysates of fibroblasts
and SHSY5Y cells. UCP2 antibody validation was initially preformed on GFP-UCP2
over-expressing SHSY5Y cells. A UCP2 immunoreactive band of approximately 120
kDa (i) was identified in the UCP2 GFP transfected cells absent in untransfected
control SHSY5Y cells (Fig 5.4A) potentially corresponding to a GFP tagged UCP2
dimer (predicted UCP monomer 33 kDa, GFP 30 kDa). In addition, a 60 kDa (ii)
immunoreactive band of four fold weaker intensity was detected in both the GFPUCP2 transfected and control SHSY5Y cells equivalent to the molecular weight of an
untagged UCP2 dimer (Fig 5.4A, B). Additional higher and lower molecular bands at
20 kDa, 55 kDa, 80 kDa and 150 kDa were also present. These bands could not
definitively be assigned to a particular UCP2 species and may represent truncated
186
endogenous protein, post-translationally modified UCP2, oligomeric forms or other
UCP isoforms as has previously been reported for this antibody [158].
The 66 kDa (ii) immunoreactive species previously assigned to the
endogenous untagged UCP2 dimer was detected in GFP UCP2, wild type and G2019S
LRRK2-V5 SHSY5Y cell extracts in similar abundance with the equivalent band 3 fold
lower in control SHSY5Y cells (Fig 5.4B). The level of the 66 kDa immunoreactive
species was three fold lower for both control and G2019S LRRK2 patient fibroblasts
relative to control SHSY5Y cells consistent with lower UCP2 mRNA expression in
these cell lines. An additional band, common to both cell types, was detected at
approximately 60 kDa (iv) with 2-4 fold higher immunoreactivity in fibroblasts
relative to SHSY5Y cells. A UCP2 immunoreactive band at 125 kDa (iii) was identified
in all fibroblast extracts potentially represented the endogenous UCP2 tetramer
[314]. In addition, fibroblast extracts were immunoreactive for 75 kDa and 90 kDa
UCP2 species, while SHSY5Y cells produced additional 70 kDa and 73 kDa UCP2
immunoreactive bands. As we were unable to definitively characterise the UCP2
species in UCP2-GFP over-expressing SHSY5Y cells it was not possible to quantify
UCP2 levels in fibroblasts and SHSY5Y cells.
187
5.1.4.1 Evaluation of the influence of genipin on UCP induced mitochondrial
depolarization in SHSY5Y cells
To further investigate the potential role of UCP protein acting downstream
of LRRK2 G2019S, the effects on mitochondrial bioenergetics by the previously
characterized UCP inhibitor, genipin were tested [315]. The optimal concentration
for this compound in SHSY5Y cells was determined by titrating increasing amounts of
genipin (1.375-5.5 nM) into control and GFP-UCP2 over-expressing cells, monitoring
the change in TMRM intensity over the course of 30 minutes (Fig 5.5A).
Untransfected control SHSY5Y cells treated with genipin (1.375 nM) showed a mild,
5 % increase in TMRM intensity potentially due to the influence on endogenous
UCPs. Genipin (1.375 nM) increased the TMRM fluorescent intensity in GFP-UCP2
cells by 20 % after 5 minutes of drug addition to levels comparable to those of
untransfected control SHSY5Y cells consistent with a rescue of the reduced
188
mitochondrial membrane potential in GFP-UCP2 cells. The increase in the TMRM
intensity was maintained for the remainder of the 30 minute time course. There was
no obvious change in the TMRM or GFP distribution with cell morphology
unperturbed at the lower dose of the drug (Fig 5.5B). GFP-UCP2 over-expressing
SHSY5Y cells treated with 2.75 nM genipin showed a linear increase in TMRM
intensity over the 30 minutes of the measurement with a 45 % increase in TMRM
fluorescence recorded at the last time point (Fig 5.5A). Higher genipin
concentrations (5.5 nM) caused a more rapid increase in TMRM intensity but made
it difficult to image the cells as the genipin treatment resulted in cell mobility and
nuclear shrinkage indicative of toxic side effects (Fig 5.5C, cells marked â–¡).
Control, wild type and G2019S LRRK2-V5 over-expressing SHSY5Y cells were
treated with 1.375 nM genipin, the concentration of the drug which restored the
mitochondrial membrane potential in UCP2-GFP expressing cells, and the
mitochondrial membrane potential monitored by TMRM staining over 20 minutes.
Consistent with previous measurements, TMRM intensity in control cells increased
by approximately 5 % (Fig 5.5D) 5 minutes following genipin (1.375 nM) treatment
with a stable TMRM signal intensity maintained for the duration of the time course.
The TMRM intensity of wild type LRRK2-V5 over-expressing cells was not affected by
genipin treatment while the TMRM intensity of G2019S LRRK2-V5 over-expressing
SHSY5Y cells gradually increased in response to genipin, with a 5-10 % increase in
TMRM fluorescence after 5 minutes and a 20 % increase after 10 minutes following
treatment relative to untreated cells consistent with a restoration of the
mitochondrial membrane potential to values resembling those observed for control
and wild type LRRK2-V5 over-expressing SHSY5Y cells (Fig 5.5D).
189
5.1.4.2 Evaluation of the influence of genipin on UCP induced mitochondrial
depolarization in fibroblasts
Two cell lines of control (c1, c7) and three G2019S (p1, p2, p7) patient
fibroblast lines were treated with genipin (1.375 nM) and the mitochondrial
190
membrane potential monitored by the intensity of TMRM staining over the course
of 30 minutes. In control cells an increase in TMRM fluorescence was recorded over
the course of the measurement with the TMRM signal measured to 110 % at 30
minutes relative to that of untreated fibroblasts (Fig 5.6A). However, 30 minutes
after genipin addition, a 20 % drop in the TMRM intensity was observed for both cell
lines accompanied by nuclear shrinkage and excessive mobility (cells marked â–¡ in Fig
5.6B) in some of the treated control fibroblasts. Three lines of G2019S LRRK2
fibroblasts showed a more pronounced increase in TMRM intensity with a linear
increase in TMRM fluorescence from the basal 80 % to 130 % for two of the G2019S
lines and to 100 % (Fig 5.6A) for the third G2019S line when compared to untreated
control fibroblasts. In contrast to the control fibroblasts, there was no detectable
TMRM intensity drop observed for the G2019S cells over the course of the
measurement and the cellular morphology remained unchanged following genipin
treatment (Fig 5.6C).
191
5.1.5 LRRK2 kinase activity assays with recombinant UCP2 protein
UCP activity has previously been shown to be regulated by phosphorylation
[159] and the mitochondrial fractionation experiments carried out in Chapter 4
identified LRRK2 associated with mitochondrial membranes. To determine whether
192
LRRK2 could mediate UCP activity directly, in vitro LRRK2 kinase activity assays were
carried out on wild type GST-LRRK2 immunoprecipitated from HEK293 cells and
recombinant GST-UCP2 protein. Recombinant proteins were generated and LRRK2
kinase assays carried out by Francisco Inesta-Vaquera (Dundee). SDS PAGE and
coomassie stain of the components of the kinase activity reaction identified wild
type GST-LRRK2 (residues 1326-2527) as the 164 kDa coomassie stained band,
absent in control reactions where LRRK2 was omitted (Fig 5.7). GST-moesin, a
previously characterised substrate for the LRRK2 kinase was detected as the 110 kDa
coomassie stained species absent in GST-UCP2 containing reactions. LRRK2 kinase
activity
reactions
contained
increasing
GST-UCP2
(2-10
µM)
substrate
concentrations, as shown by the increase in 54 kDa coomassie band intensity
corresponding to GST tagged UCP2 monomer. Additional higher and lower
molecular weight species were identified in GST-UCP2 containing reactions, absent
in the GST-Moesin lanes (*), however the identity of these species was not
investigated.
Autoradiography analysis of LRRK2 kinase activity assays containing wild
type GST-LRRK2, GST-Moesin or GST-UCP2 detected a 164 kDa band corresponding
to the truncated LRRK2 protein consistent with LRRK2 autophosphorylation. In
addition, in LRRK2 kinase activity assays containing GST-Moesin, autoradiography
detected a 110 kDa band corresponding to radiolabelled GST-Moesin consistent with
phosphorylation by GST-LRRK2. In contrast, no clear bands were detected in LRRK2
kinase activity reactions containing GST-UCP2 protein by autoradiography,
suggesting LRRK2 was not phosphorylating UCP2 in vitro under the conditions
tested.
193
5.1.6 The impact of UCP4 knockdown upon UCP4 mRNA levels and mitochondrial
membrane potential in SHSY5Y cells
To examine the role of UCP4 up-regulation in G2019S LRRK2-V5 overexpressing SHSY5Y cells we investigated the influence of UCP4 knockdown. Control
SHSY5Y cells were treated with UCP4 siRNA or scrambled siRNA for 3 days following
which mRNA was extracted, reverse transcribed to cDNA and evaluated by real time
PCR analysis. Control SHSY5Y cells treated with 10 nM and 5 nM UCP4 siRNA had 42
% and 46 % UCP4 mRNA levels relative to untreated cells when corrected for GAPDH
194
expression consistent with a marked reduction in UCP4 mRNA levels. Control SHSY5Y
cells treated with scrambled siRNA had 21 % higher UCP4 mRNA transcript levels
relative to untreated cells (Fig 5.8C).
UCP4 mRNA levels in wild type and G2019S LRRK2-V5 over-expressing
SHSY5Y cells treated with UCP4 siRNA (5 nM) were 53 % and 75 % relative to the
corresponding untreated clones consistent with a lower efficiency of UCP4 mRNA
silencing relative to that achieved in control SHSY5Y cells (Fig 5.9A).
As commercial UCP4 antibodies were not available at the time of the study
we were unable to quantify levels of UCP4 protein in UCP4 siRNA treated SHSY5Y
cells. However, UCPs have previously been reported to have relatively short half
lives [160] hence we investigated whether the reduction in UCP4 mRNA expression
was sufficient to alter the mitochondrial membrane potential in these cells.
Mitochondrial membrane potential was assessed in untreated and UCP4 siRNA (5
nM) treated control, wild type and G2019S LRRK2 V5 over-expressing SHSY5Y cells in
glucose rich respiration buffer by TMRM staining and confocal microscopy. TMRM
195
intensities of control SHSY5Y cells treated with UCP4 siRNA were comparable to
those of untreated SHSY5Y cells (Fig 5.9B). UCP4 siRNA treated wild type LRRK2-V5
SHSY5Y cells had 10 % (p<0.0001) lower TMRM fluorescence relative to untreated
cells, indicative of a reduced mitochondrial membrane potential. The reason for this
decrease is not known. G2019S LRRK2-V5 over-expressing SHSY5Y cells treated with
UCP4 siRNA had 6 % (p<0.0001) higher TMRM intensities when compared to
untreated cells, consistent with a higher mitochondrial membrane potential (Fig
5.9B) potentially due to lower UCP4 mRNA expression. However due to the low
UCP4 siRNA knockdown efficiency in wild type and G2019S expressing cells and the
effect of UCP4 siRNA treatment on the mitochondrial membrane potential in wild
type LRRK2 over-expressing SHSY5Y cells, we could not fully evaluate the effect of
UCP4 knockdown in the SHSY5Y system.
196
5.1.7 The influence of LRRK2 kinase inhibition upon UCP mRNA expression in
SHSY5Y cells
Bioenergetic assessment of wild type and G2019S LRRK2-V5 over-expressing
SHSY5Y cells in Chapter 4 indicated the mitochondrial membrane potential and
oxygen consumption abnormalities associated with G2019S expression could be
partially reversed upon LRRK2 kinase inhibition. To determine whether these
changes correlated with UCP mRNA transcript levels mRNA was extracted from
197
control, wild type and G2019S LRRK2-V5 over-expressing SHSY5Y cells following
treatment with LRRK2 IN1 (1µM, 90 min), reverse transcribed to cDNA and UCP4
mRNA expression quantified by real time PCR. Data was expressed as the
percentage change in UCP4 CT values relative to untreated control SHSY5Y cells.
UCPs 2, 4 and 5 were reduced by 15-30 % in control SHSY5Y cells treated with LRRK2
IN1 (Fig 5.10). Wild type LRRK2-V5 over-expressing SHSY5Y cells treated with LRRK2IN1 showed a 10 % decrease in UCP2 mRNA levels but expression of UCP mRNA for
isoforms 3-5 showed a considerable amount of variability between experiments
making it difficult to evaluate the effect of LRRK2-IN1 treatment on the levels of
these UCP isoforms (Fig 5.10). G2019S LRRK2-V5 expressing SHSY5Y cells treated
with LRRK2-IN1 had similar UCP abundance for isoforms 2, 3 and 5 relative to
untreated control cells. Basal UCP4 mRNA levels were 700 % higher in G2019S
LRRK2-V5 over-expressing SHSY5Y cells when compared to control cells but were
reduced to 150 % in cells treated with LRRK2 IN1 consistent with lower UCP4 mRNA
abundance following LRRK2 kinase inhibition. The reduction in UCP4 mRNA
expression in G2019S expressing SHSY5Y cells upon LRRK2 kinase inhibition
correlated with a reduced rate of mitochondrial respiration and increased
mitochondrial membrane potential (Fig 4.14 A,C).
198
5.1.8 The impact of LRRK2 knockdown upon UCP4 mRNA levels and mitochondrial
membrane potential in SHSY5Y cells
To investigate the influence of endogenous LRRK2 on the regulation of UCP
expression, UCP4 mRNA levels were evaluated in the SHSY5Y LRRK2 siRNA
knockdown model. mRNA was extracted from SHSY5Y cells treated with LRRK2
siRNA for 6 days, reverse transcribed to cDNA and assessed for UCP4 mRNA levels by
quantitative real time PCR. SHSY5Y cells treated with LRRK2 siRNA showed a 53 %
decrease in UCP4 transcript levels relative to untreated cells (Fig 5.11A) supporting a
role for endogenous LRRK2 in influencing UCP4 mRNA abundance. Mitochondrial
membrane potential was assessed in control and LRRK2 siRNA treated SHSY5Y cells
in glucose rich respiration buffer by TMRM staining and confocal microscopy. The 53
% reduction in UCP4 mRNA expression in siRNA treated SHSY5Y cells correlated with
a 10 % increase in TMRM intensity consistent a higher mitochondrial membrane
potential relative to untreated cells (Fig 5.11B).
199
5.2.1 Characterization of DJ1 knockdown SHSY5Y cells
Having shown UCP mRNA transcript levels are affected by changes in LRRK2
kinase activity as well as levels of endogenous protein, we set out to investigate
potential upstream events which may contribute to the UCP mRNA expressional
changes. DJ1 has previously been implicated in the UCP mRNA regulatory pathway in
dopaminergic neurons [168]. To determine whether DJ1 plays a role in regulating
UCP mRNA expression in SHSY5Y cells, 2 clones of SHSY5Y cells stably expressing DJ1
shRNA and 1 clone expressing scrambled shRNA were generated by Zhi Yao (UCL).
mRNA was extracted from control, DJ1 shRNA and scrambled shRNA expressing
SHSY5Y cells, reverse transcribed to cDNA and evaluated for DJ1 mRNA expression
by quantitative real time PCR analysis. DJ1 mRNA levels were similar between
control and scrambled shRNA treated SHSY5Y cells while DJ1 shRNA expression
correlated with 85 % (clone x) and 95 % (clone y) lower DJ1 mRNA transcript levels
relative to control SHSY5Y cells (Fig 5.12A). Western blot analysis of whole cell
200
lysates of control, DJ1 shRNA and scrambled shRNA expressing SHSY5Y cells showed
comparable 30 kDa DJ1 immunoreactivities in control and scrambled shRNA treated
cells (Fig 5.12 B, C) while an 80 % (clone x) and 90 % (clone y) decrease in the 30 kDa
immunoreactivity was determined for DJ1 shRNA expressing cells relative to control
cells consistent with reduced levels of DJ1 protein (Fig 5.12B, C).
5.2.2 The analysis of UCP mRNA expression in DJ1 knockdown SHSY5Y cells
To determine if DJ1 could play a role in the G2019S LRRK2 up regulation of
UCP4 in SHSY5Y cells, DJ1 knockdown cells were screened for UCP2-5 mRNA
expression. mRNA was extracted from SHSY5Y cells, reverse transcribed to cDNA and
UCP expression evaluated by quantitative real time PCR analysis. UCP2 mRNA levels
were comparable between control and DJ1 shRNA expressing SHSY5Y cells and
reduced by 40 % in the scrambled shRNA expressing cells relative to control cells,
201
although this value was not statistically significant (Fig 5.13A). UCP3 mRNA
expression was significantly lower in DJ1 shRNA expressing SHSY5Y cells (40-50 %,
p<0.05) relative to untreated cells suggesting the loss of DJ1 is affecting the
expression of the UCP3 isoform (Fig 5.13B). UCP4 mRNA expression was comparable
between control cells and DJ1 shRNA clone x while a 56 % increase in UCP4 mRNA
expression was identified for DJ1 shRNA clone y (Fig 5.13C). UCP4 mRNA levels were
147 % more abundant in scrambled shRNA cells relative to control SHSY5Y cells,
although this value was not statistically significant. UCP5 mRNA transcript levels
were comparable between control, scrambled shRNA and DJ1 shRNA clone x cells
with a 20 % reduction in UCP5 mRNA levels determined for DJ1 shRNA expressing
clone y (Fig 5.13D). Due to the lack of a clear role for DJ1 in the regulation of UCP4
mRNA levels we did not pursue with knocking DJ1 down in G2019S LRRK2-V5
expressing SHSY5Y cells.
202
5.3 The influence of LRRK2 expression upon PGC1α, SOD2 and catalase mRNA in
SHSY5Y cells
UCP mRNA expression has previously been shown to be regulated by
PGC1α/PPARγ mediated transcriptional activation by increased PGC1α levels [316].
PGC1α mRNA expression was quantified by real time PCR in control, wild type and
G2019S LRRK2-V5 over-expressing SHSY5Y cells. PGC1α mRNA levels were
comparable between control and wild type LRRK2-V5 over-expressing SHSY5Y cells
while a 70 % increase in PGC1α mRNA was identified for the G2019S LRRK2-V5 cells
consistent with increased PGC1α expression (Fig 5.14A). As PGC1α levels have also
been shown to regulate expression of antioxidant defence genes including SOD2 and
203
catalase, we evaluated the mRNA levels of SOD2 and catalase in these cells. SOD2
(Fig 5.14C) and catalase (Fig 5.14B) mRNA levels were comparable between the
control, wild type and G2019S LRRK2-V5 over-expressing cells.
5.4.1 The analysis of HDAC5 cellular distribution by immunohistochemistry in
SHSY5Y cells
HDAC5 involvement in PGC1α regulation was investigated by measuring
nuclear and cytosolic HDAC5 protein by immunocytochemistry. Control SHSY5Y cells
were fixed in paraformaldehyde and permeabilised in 0.1 % Triton X100. Titrating
HDAC5 antibody concentrations (50-200 nM) revealed a dose dependent increase in
the green fluorescence in SHSY5Y cells (Fig 5.15A). This increase in fluorescence was
observed for large nuclear punctate structures (nuclear HDAC5 pool), diffuse
cytosolic staining (cytoplasmic HDAC5 pool) and punctate cytosolic structures
potentially representing HDAC5 polymers (complexes). Concentrations (200 µM)
204
where the nuclear HDAC5 punctate structures were best resolved were chosen for
subsequent experiments.
Phorbol myristate acetate (PMA) has previously been shown to induce HDAC
nuclear export [317] and was used to test the HDAC5 antibody specificity in SHSY5Y
cells monitoring the nuclear to cytosolic HDAC5 redistribution. The PMA dose (100
nM) was chosen based on concentrations previously shown to induce HDAC5
nuclear export in SHSY5Y cells [317]. Control SHSY5Y cells treated with PMA (100
nM, 90 min) resulted in a decrease in HDAC5 immunofluorescent intensity of nuclear
immunoreactive punctae with a weak increase in cytosolic HDAC5 immunoreactivity
consistent with a partial redistribution of the HDAC5 molecule from the nucleus into
the cytosolic compartment. Cells subject to prolonged PMA treatment (3 hours)
showed a further reduction in the intensity of the nuclear HDAC5 signal with the
nuclear punctate structures appearing less well defined to the ones observed in the
untreated SHSY5Y cells (Fig 5.15B). The intensity and frequency of the cytosolic
punctate structures were comparable between untreated and PMA treated cells. As
we did not observe gross changes in cellular HDAC5 distribution in response to PMA
we did not extend the immunohistochemical analysis to wild type and G2019S
LRRK2-V5 over-expressing SHSY5Y cells.
205
5.4.2 The analysis of HDAC5 by Western blot in SHSY5Y cells
HDAC5 protein levels were evaluated in SHSY5Y cells by Western blot
analysis of whole cell lysates. The HDAC5 antibody showed strong immunoreactivity
for the 100 kDa species on Western blot of control SHSY5Y cells corresponding to
206
the predicted full length HDAC5 protein (Fig 5.16). Comparable 100 kDa HDAC5
immunoreactivities were identified following PMA treatment (100 nM, 90 min)
suggesting PMA does not influence cellular HDAC5 levels (Fig 5.16). Comparison of
HDAC5 immunoreactivity between control, wild type and G2019S LRRK2-V5 overexpressing SHSY5Y cells revealed similar HDAC5 band intensities relative to GAPDH
immunoreactivity correcting for equivalent protein loading, in the absence and
presence of PMA consistent with equivalent levels of cellular HDAC5 protein.
5.4.3 The analysis of HDAC5 and LRRK2 interaction
HDAC5 immunoprecipitations were carried out on 1 % Triton X100 extracts
of LRRK2 over-expressing SHSY5Y cells, control and LRRK2 knockout mouse brain
cortex to investigate whether LRRK2 interacts with HDAC5. In addition, HDAC5
immunoprecipitations were carried out on LRRK2 over-expressing SHSY5Y cells
treated with PMA in an attempt to enhance the cytosolic HDAC5 levels. Western
blot analysis of HDAC5 immunoprecipitates from wild type and LRRK2 knockout
mouse brains detected a HDAC5 immunoreactive species of approximately 100 kDa
in all brain IPs, absent in IPs where HDAC5 antibodies were omitted (Fig 5.17A). The
207
100 kDa band co-migrated with the HDAC5 immunoreactivity detected in brain IP
input lysate consistent with a successful pull-down of a full length HDAC5. Western
blot analysis of HDAC immunoprecipitates revealed 3514-1 weak 280 kDa
immunoreactivity detected in the wild type brains absent in LRRK2 knockout brain
samples. 3514-1 280 kDa immunoreactivity could not be detected in the brain IP
input lysate due to low LRRK2 levels, as has previously been shown in Chapter 3. An
additional band of 170 kDa (*) was detected by the 3514-1 antibody in HDAC5 IP
from mouse brains as well as a weaker intensity signal in the input fraction. As the
signal was present in both control and LRRK2 knockout brains it was possible to
assume non specific binding of the antibody. The co-IP of LRRK2 with HDAC5
suggested LRRK2 interacted with full length HDAC5 in the mouse brain.
To determine whether the LRRK2 HDAC5 association occurs in SHSY5Y cells,
HDAC5 was immunoprecipitated from 1 % Triton X100 extracts of control, wild type
and G2019S LRRK2-V5 over-expressing SHSY5Y cells. Positive immunoreactivity for
the HDAC5 antibody was detected in the IP input sample of SHSY5Y cells, consistent
with full length HDAC5 protein (Fig 5.17B). Equivalent HDAC5 100 kDa
immunoreactive bands were detected in immunoprecipitates from control, wild type
and G2019S LRRK2-V5 over-expressing SHSY5Y cells consistent with comparable
efficiencies in the HDAC5 IP (Fig 5.17B). An additional lower molecular weight
species immunoreactive for HDAC5 was detected at approximately 90-95 kDa in the
IP samples, absent in the IP input lysate (Fig 5.17B), control and LRRK2 knockout
mouse brains (Fig 5.17A). The nature of this species was not evaluated. Probing the
IP Western blot for V5 immunoreactivity detected a weak immunoreactive band in
the wild type LRRK2-V5 over-expressing samples and a 280 kDa species of
approximately two fold greater intensity in the G2019S LRRK2-V5 co-IP. No
208
equivalent bands were detected in untransfected control SHSY5Y cells or PMA
treated samples suggesting PMA influenced the LRRK2 HDAC5 association.
The G2019S LRRK2-V5 protein appeared to be more associated with HDAC5
when compared to wild type LRRK2 as shown by the increase in the V5 band
intensity of HDAC5 co-IPs (Fig 5.17B), suggesting LRRK2 kinase activity may play a
role in HDAC5 association. To investigate the effects of LRRK2 kinase inhibition on
the LRRK2 HDAC5 co-IP, immunoprecipitation was carried out following treatment of
wild type and G2019S LRRK2-V5 over-expressing SHSY5Y cells with LRRK2 kinase
inhibitors (LRRK2-IN1 (1 µM, 90 min) and CZC25146 (0.3 µM, 60 min)). LRRK2 kinase
inhibition by LRRK2-IN1 and CZC25146 and potentially PMA was evaluated by
immunoprecipitating LRRK2 from 1 % Triton X100 lysates and probing for
phosphorylation at Serine 935 by Western blot. LRRK2 Serine 935 phosphorylation
was reduced by 80-90 % in both wild type and G2019S LRRK2-V5 over-expressing
SHSY5Y cells following LRRK2-IN1 and CZC25146 treatment as determined by the
band intensity of phospho Serine 935 relative to total LRRK2 (3514-1 antibody)
consistent with the inhibition of the LRRK2 kinase activity (Fig 5.17C). In contrast, the
phospho Serine 935 immunoreactive LRRK2 band from PMA treated wild type and
G2019S LRRK2-V5 SHSY5Y cells was of comparable intensity relative to untreated
cells suggesting PMA did not influence the LRRK2 Serine 935 phosphorylation status
(Fig 5.17C).
HDAC5 was immunoprecipitated from 1 % Triton X100 lysates of untreated,
LRRK2 IN1 or CZC25146 treated wild type or G2019S LRRK2-V5 over-expressing
SHSY5Y cells. Western blot analysis of HDAC5 immunoprecipitates revealed
comparable immunoreactivities in the presence and absence of LRRK2 kinase
inhibition consistent with equivalent HDAC5 IP efficiencies (Fig 5.17D). Probing
HDAC5 IP Western blots for V5 immunoreactivity revealed an immunoreactive band
209
of approximately 2 fold greater intensity for the untreated G2019S LRRK2-V5 cell
line relative to untreated wild type LRRK2-V5 SHSY5Y cells, in line with the previous
experiment (Fig 5.17B). The 280 kDa V5 immunoreactive signal was of comparable
intensity for wild type and G2019S LRRK2-V5 over-expressing SHSY5Y cells treated
with both LRRK2-IN1 and CZC25146 LRRK2 kinase inhibitors relative to untreated
cells suggesting LRRK2 kinase inhibition does not influence the LRRK2 HDAC5
interaction (Fig 5.17D) as detected by co-IP. An additional 300 kDa band
immunoreactive for the 3514-1 antibody (*) was detected in the IP input lysates of
both wild type and G2019S LRRK2 over-expressing SHSY5Y cells, the identity of this
band was not investigated.
210
5.5.1 The analysis of DLP1 protein in SHSY5Y cells
A LRRK2 DLP1 interaction has recently been described [280, 281] correlating
with changes in mitochondrial function described for wild type and G2019S
expressing SHSY5Y cells. To evaluate the potential role of DLP1 in the regulation of
211
mitochondrial uncoupling in G2019S expressing fibroblasts and SHSY5Y cells, DLP1
levels were assessed in whole cell lysates of control and G2019S LRRK2 fibroblasts by
Western blot analysis. 90 kDa DLP1 immunoreactivity was detected in fibroblast
extracts corresponding to the predicted molecular weight of the monomeric full
length DLP1 (Fig 5.18A). Quantification of the DLP1 immunoreactivity relative to
GAPDH confirmed equivalent levels of DLP1 abundance in control and G2019S
fibroblasts (Fig 5.18B). Furthermore, DLP1 protein levels were also comparable
between control, wild type and G2019S LRRK2-V5 over-expressing SHSY5Y cells as
determined by DLP1 90 kDa immunoreactivity relative to GAPDH in whole cell
extracts (Fig 5.18C).
Mitochondrial levels of DLP1 were evaluated by Western blot analysis of
affinity purified mitochondria isolated from wild type and G2019S LRRK2-V5 overexpressing SHSY5Y cells. The band intensity of Core Complex III markers was
comparable for wild type and G2019S LRRK2-V5 mitochondrial isolations suggesting
similar mitochondrial loading (Fig 5.18D). DLP1 90 kDa immunoreactive bands in
extracts of purified mitochondria were of similar intensities for wild type and
G2019S LRRK2-V5 expressing cells with comparable DLP1/Core Complex III ratios
suggesting similar mitochondrial DLP1 levels. Furthermore, affinity purified
mitochondria from wild type and G2019S LRRK2-V5 over-expressing cells treated
with LRRK2-IN1 (1 µM, 90 min) revealed comparable DLP1 Core Complex III
immunoreactivities. Although mitochondrial DLP1 quantifications were only carried
out once (n=1) our data suggested LRRK2 G2019S and LRRK2 kinase inhibition did
not impact on steady state mitochondrial DLP1 abundance (Fig 5.18E).
212
5.5.2 The analysis of LRRK2 and DLP1 interaction in SHSY5Y cells
To determine whether LRRK2 interacts with DLP1 in SHSY5Y cells LRRK2 was
immunoprecipitated from 1 % Triton X100 extracts of wild type and G2019S LRRK2V5 over-expressing SHSY5Y cells. Western blot analysis of immunoprecipitates
revealed a 3514-1 immunoreactive 280 kDa band corresponding to the approximate
molecular weight of the full length LRRK2 protein, co-migrating with the 280 kDa
213
immunoreactive species in the input fraction and absent in immunoprecipitates of
samples where the primary antibody was omitted (Fig 5.19). Western blot analysis
of LRRK2 immunoprecipitates detected a 90 kDa DLP1 immunoreactive band
corresponding to the full length protein in the input sample, representing
approximately 5 % of the total IP input. DLP1 immunoreactivity was not clearly
detected in immunoprecipitates where the LRRK2 antibody was omitted, control,
wild type or G2019S LRRK2-V5 IP samples at the equivalent exposure times.
However, longer exposures detected a faint 90 kDa immunoreactive band
corresponding to the molecular weight of the full length DLP1 protein in
immunoprecipitates of G2019S LRRK2-V5 expressing SHSY5Y cells absent in control,
wild type LRRK2-V5 immunoprecipitates and IPs where DLP1 antibodies were
omitted consistent with a small portion of the LRRK2 G2019S mutant pool
interacting with the DLP1 molecule.
214
215
Chapter 6: Discussion
6.1 LRRK2 tissue distribution
Idiopathic PD is characterized by the loss of dopaminergic neurons in the
SNpc while the Braak hypothesis of the pathological spread in PD suggests
progressive disease transmission from the lower brain stem regions, midbrain and to
the neocortex. The motor dysfunction and cognitive decline associated with the
deterioration of these tissues suggests the function of neurons in these brain regions
may play an important role in the disease pathogenesis. As LRRK2 mutations are
clinically and pathologically indistinguishable from idiopathic PD, LRRK2 tissue
distribution may provide an insight into the function of the protein in the context of
PD pathology.
Previous studies have demonstrated ubiquitous LRRK2 expression, with low
levels of LRRK2 mRNA and protein detected in the rodent brain relative to peripheral
tissues such as the kidneys lungs and lymph nodes [245]. High LRRK2 expression in
peripheral tissues has been linked to cell specific processes such as microbial
elimination in bone derived macrophages and RAW264.7 cell lines [241, 243],
antibody secretion in B lymphocytes [242] and autophagic control of lung and kidney
homeostasis [238, 240]. Although similar processes may be relevant in the brain
potentially playing a role in PD pathogenesis, lower levels of LRRK2 mRNA imply
neurons are less dependent on these LRRK2 regulated mechanisms, supported by a
lack of a neuronal phenotype in LRRK2 knockout animals where kidney atrophy and
poor immune responses have been noted [238, 240, 271]. Furthermore, the G2019S
mutation does not appear to have an aberrant effect on the LRRK2 functions
described in peripheral tissues. The immune response in T cells [271] and microbial
elimination by macrophages [241] were unaffected by G2019S transgene expression.
Mice expressing bacterial artificial chromosome (BAC) derived wild type and G2019S
216
LRRK2 constructs had normal kidney and lung function [238] suggesting the
mutation does not functionally impair these cell types.
Proposed brain specific functions for LRRK2 include maintenance of
dopamine homeostasis which relies on the function of dopaminergic cells while the
dysregulation of this pathway is linked to PD pathogenesis [234]. Striatal dopamine
release in mice was enhanced by wild type LRRK2 over-expression and reduced in
G2019S transgenic animals, however, dopaminergic cell loss was not observed
questioning the suitability of the LRRK2 mouse model in the study of PD [234].
Although LRRK2 expression has been noted in dopaminergic brain areas, with
immunohistochemical analysis detecting LRRK2 in the striatum, hippocampus and
cortex of rodent and non human primates, substantially lower LRRK2 protein levels
were detected in the substantia nigra [244, 249, 250, 318]. The G2019S mutation
appears to be linked to the death of dopaminergic neurons in substantia nigra of PD
patients which seems to correlate with low SNpc LRRK2 expression suggesting a
specific functional role for LRRK2 in this cell type or the preferential vulnerability of
these cells to a G2019S LRRK2 associated insult.
Our analysis of LRRK2 mouse, rat and marmoset brain distribution by the
combined immunoprecipitation and Western blot approach supported previous
reports describing the LRRK2 brain expression pattern, detecting greatest LRRK2
levels in the striatum, cortex and cerebellum with lower LRRK2 abundance in the
midbrain, olfactory lobe and brain stem. Comparing LRRK2 striatal and cortical
expression to that of lymphoblasts detected 50 fold higher LRRK2 levels in the
immune cells in support of the expressional differences between peripheral cells and
brain tissue.
217
6.2 LRRK2 detection
6.2.1 LRRK2 detection by Immunohistochemistry
Methods for LRRK2 protein detection in rodent and primate brains have
previously included Western blotting, BN PAGE and immunohistochemistry using in
house and commercially available polyclonal LRRK2 antibodies. Although generally a
common distribution pattern has been noted (described in section 6.1),
discrepancies between groups exist, most pronounced in immunohistochemical
studies where LRRK2 expression in neuronal sub populations and glial cells has been
debated. In particular immunohistochemical analysis of LRRK2 expression in Purkinje
cerebellar granular cells have yielded mixed reports in both human and rodent
tissues. Three studies reported positive LRRK2 immunoreactivity [206, 319, 320]
while two others failed to detect LRRK2 in this cell subtype [246, 247]. Our
characterization of the 3514-1 antibody in wild type mouse cerebellar sections
revealed positive staining in this cell subtype absent in the equivalent LRRK2
knockout brain regions. Utilizing the same protocol we were able to detect LRRK2
immunoreactivity in human purkinje neurons of the cerebellum suggesting LRRK2 is
indeed present in this cell type in the human brain. LRRK2 expression in brain
microglia has also been debated with some reports suggesting glial cells do not
express LRRK2 [249] while other studies using the 3514-1 and NB300-268 antibodies
detected strong immunoreactivity in activated mouse microglia, human microglia
and astrocytes [320, 321]. LRRK2 glial staining was not observed in control mouse or
human cerebellar sections described here potentially due to the young age of the
animals and absence of immunoinfiltration in the human cerebellum, a point that
was not addressed. Considering the emerging literature on the role of LRRK2 in
activated macrophages [241] and adhesive monocytes [322], it is possible that
218
LRRK2 does indeed have a role in brain immunoprotection which may be relevant to
PD.
Although immunohistochemical assessment can distinguish between cell
populations in different brain regions, this technique is semi quantitative and cannot
be used to compare LRRK2 brain expression between cell types, brain regions and
species. In addition, although the specificity of the antibody can be validated by
LRRK2 knockout tissue, the size of the immunoreactive LRRK2 species detected
cannot be defined. Western blotting on the other hand is quantitative and provides
molecular weight estimations for the immunoreactive species.
6.2.2 LRRK2 detection by Western blotting
Western blot analysis for LRRK2 detection has been limited by the
availability of antibodies that can detect low levels of endogenous LRRK2 protein
and poor characterization of the <280 kDa LRRK2 immunoreactive species detected
by these antibodies. Biskup et al, [237] [251] screened 8 in house and commercially
available LRRK2 antibodies in SHSY5Y cells, M17, PC12, HEK293 cells and brain
tissue. All of the antibodies detected multiple bands in both wild type and LRRK2
knockout mouse brains including a 280 kDa band corresponding to the full length
LRRK2. The lack of specificity questioned the use of these antibodies [237, 251]. 280
kDa LRRK2 immunoreactivity in HEK293 cells was observed for two of the 8
antibodies and although the data was not shown, the authors stated that they had
difficulty detecting endogenous LRRK2 in SHSY5Y, M17 and PC12 cells. The antibody
screen described in this report detected 280 kDa LRRK2 immunoreactive species by
Western blot in SHSY5Y cells over-expressing LRRK2 absent in control cells
corresponding to the predicted molecular weight of the full length protein for all
four antibodies analyzed. However, the four antibodies could not definitively
identify endogenous 280 kDa LRRK2 in SHSY5Y cells, fibroblasts or brain extracts. In
219
addition to the 280 kDa LRRK2 immunoreactive species, lower molecular weight
immunoreactive bands were also detected in all cells and tissues analysed in line
with previous observations [237]. Similar to the <280 kDa LRRK2 immunoreactive
species in the Biskup et al report, these bands could represent non specific binding
of the antibodies however truncated versions of the protein and splice variants of
endogenous LRRK2 cannot be ruled out.
A number of studies validated in house antibodies for detection of
endogenous full length LRRK2 species by siRNA knockdown of the protein in cell
culture systems [224, 239, 243]. siRNA knockdown of LRRK2 confirmed endogenous
LRRK2 detection in Hela cells [239] and Raw264.7 macrophages [243] potentially
reflecting higher LRRK2 abundance in these cell types or improved antibody
specificity for conditions used in the study involving HEK293 cells [224]. In all cases,
the intensity of the 280 kDa band was reduced in siRNA treated cells consistent with
reduced LRRK2 abundance. However, the effect of siRNA on the band intensity of
<280 kDa immunoreactive LRRK2 species was not provided with the published data
making it difficult to address the nature of these species.
Analysis of the LRRK2 coding sequence derived from human SNpc predicted
the 280 kDa size of the multidomain protein. When cloned in carcinoma cells, the
LRRK2 immunoreactive band resolved at 280 kDa on SDS PAGE consistent with the
expression of full length monomeric LRRK2 [203]. As cDNA was cloned into the
HEK293 expression system, potential LRRK2 splice isoforms were not investigated.
The possibility of tissue specific regulation required to initiate transcription or
generate spliced and truncated forms of LRRK2 cannot be ruled out. In particular,
BAC LRRK2 derived transcripts (containing the complete LRRK2 gene) appear to
undergo post transcriptional or post-translational processing in mouse kidneys to
generate a lower molecular weight protein observed by SDS PAGE absent in
220
equivalent extracts of brain or lung tissues and LRRK2 knockout controls [238].
Characterization by mass spectrometry to identify the exact nature of the lower
molecular weight species confirmed the 160 kDa band as N terminally truncated
LRRK2 [238]. Western blot analysis of LRRK2 protein expression described in this
thesis revealed 160 kDa immunoreactivity for the 100-500 and 3514-1 antibodies in
lymphoblasts and human brain tissue. As the 100-500 antibody was raised against
the LRRK2 N terminus, the 160 kDa N terminally truncated LRRK2 would not bind
this antibody, hence the 100-500 160 kDa immunoreactive bands are likely to be
non specific. However, the 3514-1 antibody epitope spans the core domain of LRRK2
suggesting the 160 kDa LRRK2 and 3514-1 160 kDa immunoreactivity in
lymphoblasts and human brain tissue could represent the LRRK2 N terminally
truncated species. In addition, 160 kDa immunoreactivity was identified by 100-500
and 3514-1 antibodies in SHSY5Y cells over-expressing LRRK2, absent in control cells.
Alternative LRRK2 splicing was ruled out in this cDNA expressing system, however
post translational cleavage of full length LRRK2 may account for these
immunoreactive species. LRRK2 Immunoreactive bands >280 kDa have been
assigned to SDS stable dimers [244]. These > 280 kDa species were not detected in
our analyses and were absent in a range of other antibody screens [237, 249] under
a variety of solubilisation conditions tested questioning the reproducibility of these
findings. Although phosphorylation of LRRK2 has been confirmed by mass
spectrometry the species to size correlation has not yet been fully clarified [202,
211].
6.2.3 LRRK2 detection by BN PAGE
LRRK2 detection by BN PAGE has previously been used to look at the native
structure of the protein. In lymphoblasts and HEK293 cells over-expressing LRRK2,
LRRK2 was shown to exist predominantly as a dimer and in higher molecular weight
221
oligomeric forms while monomeric LRRK2 could not be detected in these systems
under native conditions [228, 323]. The lack of native monomeric LRRK2 was
challenged by Berger et al, [252] who demonstrated the 280 kDa as the predominant
native species in mouse brain, lymphoblasts and HEK293 over-expressing cells. We
investigated the native state of LRRK2 and the specificity of the 3514-1 LRRK2
antibody in over-expressing SHSY5Y cells and endogenous brain tissue using this
methodology.
LRRK2 V5 immunoreactivity confirmed the equivalent species detected by
the 3514-1 antibody represented over-expressed LRRK2 in SHSY5Y cells. A 280 kDa
species was the dominant form of LRRK2 detected in over-expressing SHSY5Y cells,
human and rodent brains by BN PAGE electrophoresis in line with reports by Berger
et al. This may in part be due to the preferential extraction of the soluble pool of
LRRK2 using the extraction conditions for BN PAGE, as LRRK2 dimers have previously
been shown to be enriched in the membrane fraction [252]. In support of this,
inclusion of DDM and Triton X100 detergents during solubilisation of SHSY5Y cells
over-expressing LRRK2 increased the recovery of the higher molecular weight LRRK2
species. Although detergents appeared to improve recovery of LRRK2, the resolution
and LRRK2 migration on BN PAGE resulted in the formation of immunoreactive
smears, previously assigned to higher molecular weight LRRK2 species [228, 323].
Both the V5 and 3514-1 antibodies showed comparable specificities for the
monomeric and oligomeric forms of LRRK2 in contrast to SDS PAGE and Western
blot analysis where multiple lower molecular weight species were detected by 35141, potentially due to improved specificity of the 3514-1 antibody for LRRK2 under
native conditions.
Assessment of human and rodent samples by BN PAGE produced a range of
LRRK2 immunoreactive species not resolved in the over-expressing SHSY5Y cells
222
potentially due to low levels of tissue soluble LRRK2, supported by the higher ratio
of membrane versus soluble LRRK2 in rodent tissue relative to lymphoblasts and
LRRK2 over-expressing HEK293 cells [252]. In addition, a 280 kDa immunoreactive
band was detected in LRRK2 knockout brain extracts suggesting non specific binding
of the antibody may occur under these solubilisation conditions. As mouse LRRK2
knockout tissue has not previously been evaluated by BN PAGE, the true nature of
the bands observed in lymphoblasts and brain tissue warrants further analysis. The
poor recovery and resolution on Western blot of the LRRK2 immunoreactive species
in mouse and human brain extracts did not allow for quantitative comparison of
LRRK2 protein levels between different regions of human and rodent brain samples.
6.2.4 LRRK2 detection by immunoprecipitation and Western blotting
The
immunoprecipitation
protocol
utilized
throughout
this
study
demonstrated specificity and improved sensitivity of antibody detection when
compared to Western blot analysis and BN PAGE. The combination of two LRRK2
antibodies consistently produced a 280 kDa species in all cells, rodent and non
human primate brain tissue. The LRRK2 immunoreactive band could not be detected
in knockout mouse brain tissue and the signal intensity was markedly reduced in
SHSY5Y cells upon siRNA treatment. Solubilization conditions producing efficient >90
% LRRK2 recovery and the linear correlation for the LRRK2 signal intensity change
relative to protein input made it comparable to the quantitative assessment of
whole cell lysates by SDS PAGE [244]. IP assessment suggested the 280 kDa species
was the dominant isoform in all models analyzed and that the truncated LRRK2
protein previously validated by mass spectrometry [238] was not detected by this
methodology. As the 100-500 antibody was raised against the N terminus of LRRK2,
N terminally truncated 160 kDa LRRK2 would not bind the 100-500 antibody. The
absence of additional bands implied enhanced specificity for the 250-300 kDa
223
protein or reduced sensitivity for the lower molecular weight species suggesting the
combination of IP and Western blot preferentially detects the full length LRRK2.
In contrast to the immunohistochemical analyses where LRRK2 was readily
detected in both human and mouse cerebellum [248, 251, 319, 320],
immunoreactivity was either absent or extremely weak in Western blot analysis of
immunoprecipitates from human brain lysates of the frontal cortex, amygdala,
caudate nucleus and cerebellum. Even when analysing samples with post mortem
delay times as short as 1h30. Due to the lack of a suitable control for the human
tissue,
the
weak
intensity
identified
as
LRRK2
immunoreactivity
in
immunoprecipitates from one human cerebellum extract cannot be definitively
confirmed as a true LRRK2 species, an issue which has been encountered in other
studies [318]. Analysis of LRRK2 recovery from fibroblast lysates in the presence of
human brain samples revealed a reduction in the signal intensity suggesting the
human brain sample was interfering with the IP procedure. This phenomenon was
not observed in the presence of mouse brain extracts suggesting quantifications of
LRRK2 distribution in the rodent brain are not underestimated and the problem with
LRRK2 detection in immunoprecipitates from human brain extracts is likely to be
influenced by the contents of the human brain lysate.
Based on our evaluation of immunohistochemistry, Western blotting, BN
PAGE and immunoprecipitation in combination with Western blot analysis as
methods for LRRK2 detection and quantification of protein expression, the
combined immunoprecipitation and Western blot approach was chosen for
quantitative comparison of the relative LRRK2 expression in fibroblast, lymphoblast,
SHSY5Y cell lines and dopaminergic mouse brain areas in an attempt to evaluate the
suitability of our cell systems as models to study LRRK2 function and the effects of
the G2019S mutation in relation to PD pathogenesis.
224
6.3 Evaluation of LRRK2 protein expression in cell models and brain tissue
LRRK2 protein was detected in fibroblasts, lymphoblasts, SHSY5Y cells and
mouse brain tissue with the greatest expression reported in SHSY5Y LRRK2 overexpressing cells. Although fibroblasts have previously been shown to express LRRK2
mRNA [324], LRRK2 protein expression in this cell type was not demonstrated. Using
the combined immunoprecipitation and Western blot approach, LRRK2 protein was
detected in control human fibroblasts at 10 and 20 fold lower abundances relative
to lymphoblasts and over-expressing SHSY5Y cells respectively. LRRK2 expression in
fibroblasts was quantified as 10 fold higher relative to the striatal and cortical mouse
brain regions as well as endogenous protein in SHSY5Y cells. Considering fibroblasts
express low levels of endogenous LRRK2 and the availability of 8 cell lines from PD
patients harbouring the G2019S mutation this cell model was determined as an
attractive system to study the generic cellular consequences of G2019S LRRK2 in
relation to PD pathogenesis.
Lymphoblasts and B lymphocytes, cells from which the lymphoblasts were
derived have previously been shown to express high levels of endogenous LRRK2
protein [242, 244] detectable by Western blot analysis of whole cell lysates.
However, LRRK2 expression was reported as variable when 5 lymphoblast cell lines
were evaluated [228]. We detected endogenous LRRK2 in lymphoblasts by Western
blot (NT2, NB300-268 antibodies) and confirmed LRRK2 expression by the combined
immunoprecipitation and Western blot approach. Western blot analysis detected
variability in LRRK2 expression in 6 lines from control and 8 lines from G2019S LRRK2
PD patients, in line with previous observations [228]. Endogenous expression of
LRRK2 protein in the highest expressing lymphoblast cell line was comparable to
LRRK2 levels detected in primary PMBCs suggesting LRRK2 expression in
lymphoblasts was physiologically relevant. PMBC fractionation revealed high LRRK2
225
expression in B cells and lower abundance in the T cell sub-types consistent with
previous observations [242, 325]. Although LRRK2 protein levels were comparable
between lymphoblasts and primary PMBCs, the variability in LRRK2 levels between
cell lines would complicate the interpretation of the effect of LRRK2 function in
relation to PD. However, the ability to detect endogenous LRRK2 by Western blot in
lymphoblasts prompted us to study endogenous LRRK2 sub-cellular distribution in
this cell type.
SHSY5Y cells were chosen as a neuronal cell model to investigate LRRK2
function. Although endogenous LRRK2 expression in SHSY5Y cells was similar to
protein levels detected in mouse striatum and cortex, we were limited to
endogenous wild type LRRK2 expression in this cell type and were unable to
investigate the effects of endogenous G2019S LRRK2. Hence SHSY5Y cells were
utilized as a system to look at LRRK2 knockdown. LRRK2 has previously been shown
to have a long half life 24-36 hours [201] in HEK293 cells and 3 day LRRK2 siRNA
silencing of endogenous protein in RAW264.7 cells reduced the levels of the 280 kDa
species by >90 % [241]. SHSY5Y cells described here treated with LRRK2 siRNA for a
period of 3 days were shown to have 60 % less endogenous LRRK2 protein as
determined by the combined immunoprecipitation and Western blot approach.
Extending the LRRK2 siRNA treatment to 6 days achieved an 85 % reduction in
endogenous LRRK2 protein in SHSY5Y cells. Although previous reports [257, 326,
327] have used 3 day silencing protocols, the greater knockdown efficiency following
6 days of LRRK2 silencing convinced us to utilize the 6 day silencing protocol to
address the functional relevance of the loss of LRRK2 protein in the context of
molecular pathways implicated in Parkinson’s disease.
The LRRK2 over-expressing SHSY5Y model was generated to address the
effects of the G2019S mutation on LRRK2 function in a neuronal system. The relative
226
level of LRRK2 expression in the over-expressing SHSY5Y cells was 2-3 fold greater
than levels detected in the highest expressing lymphoblast lines. Considering LRRK2
protein levels in RAW624.7 cells and B lymphocytes have been shown to respond to
IFNγ and LPS rising to 10 fold levels of magnitude relative to basal LRRK2 expression
[241, 243], the level of LRRK2 over-expression in SHSY5Y cells can be considered high
but within the physiological range.
Using the combined immunoprecipitation and Western blot approach
quantification of endogenous LRRK2 expression in control SHSY5Y cells and rodent
striatum determined wild type LRRK2 over-expression in SHSY5Y cells to be 100 fold
greater relative to endogenous levels of LRRK2 in the striatum and control SHSY5Y
cells. Wild type and mutant LRRK2 over-expression has previously been associated
with induction of apoptosis in HEK293 and HeLa cells [254]. Furthermore, overexpression of both the wild type and mutant LRRK2 protein has previously been
shown to compromise cell viability in SHSY5Y cells [227] suggesting increased levels
of LRRK2 are potentially toxic and questioning the suitability of a LRRK2 overexpressing cell model system. Although dopaminergic cell death is observed in SNpc
of PD patients there is no evidence to suggest LRRK2 levels are altered in the PD
brain [328]. Furthermore there was no evidence of cell death in brains or peripheral
tissues of wild type and mutant LRRK2 transgenic mice suggesting LRRK2 overexpression in animal tissues does not result in apoptosis [240, 329, 330]. In addition,
induction of LRRK2 protein expression in peripheral tissues leading to increased
LRRK2 protein levels have been linked to functional roles of microbial elimination
[243], T cell activation and cytokine secretion [271] with no correlation to apoptosis.
Cell viability was not directly assessed in the SHSY5Y over-expressing cells
however there was no obvious evidence of cell death or differences in growth rates
between control, wild type and G2019S LRRK2 over-expressing cells. The
227
discrepancy between the SHSY5Y system described here and by Iaccarino et al may
be due to the differences in the level of over-expression. LRRK2 over-expression in
SHSY5Y cells described by Iaccarino et al was under the control of a CMV promoter
potentially driving higher levels of over-expression.
As the SHSY5Y cells over-
expressing LRRK2 presented here have comparable LRRK2 protein levels to those
detected in lymphoblasts, greater LRRK2 expression may be required to induce
toxicity.
A number of studies have reported cellular phenotypes associated with
induced or over-expressed wild type LRRK2 protein suggesting functions governed
by increased LRRK2 levels which may be independent of G2019S associated
processes and relevant to the SHSY5Y LRRK2 over-expressing model. Microbial
elimination by bone marrow derived macrophages involved the up-regulation of
endogenous wild type LRRK2 protein [241]. Synaptic vesicle endocytosis was
impaired by wild type LRRK2 over-expression and LRRK2 siRNA knockdown in
primary neuronal cultures [255]. LRRK2 over-expression appeared to deplete the
soluble β-tubulin pool in wild type and G2019S transgenic mouse brains consistent
with enhanced tubulin polymerization [262]. To help identify artefacts associated
with LRRK2 over-expression, the SHSY5Y over-expressing model and fibroblasts
expressing endogenous wild type and G2019S LRRK2 were assessed in parallel.
6.4 Evaluation of LRRK2 G2019S protein and mRNA expression
As previously described (section 6.3) LRRK2 function can be regulated by
induction of expression by LPS, IFNγ and microbial infiltration with subsequent
changes in steady state LRRK2 mRNA and protein levels. We set out to determine
whether the G2019S mutation correlated with altered LRRK2 expression. LRRK2
mRNA levels have previously been assessed as part of a microarray analysis carried
out on the same 6 control and 8 G2019S fibroblast cell lines used in this study,
228
finding comparable LRRK2 mRNA levels in control and G2019S cells [324].
Quantitative real time PCR analysis described here confirmed comparable levels of
endogenous LRRK2 mRNA for wild type and mutant fibroblast cell lines. In addition,
similar levels of LRRK2 mRNA have been described for control and G2019S LRRK2 PD
brains as noted in the amygdala, putamen, cerebellum and frontal cortex regions
further suggesting the pathogenic mutation does not correlate with differential
LRRK2 mRNA expression [331].
LRRK2 polymorphisms have previously been shown to affect the stability of
the LRRK2 protein. The T2397M polymorphism prevalent in patients suffering from
inflammatory bowel disease has been demonstrated to result in lower levels of
LRRK2 protein in B lymphocytes without affecting LRRK2 mRNA expression [271]
suggesting single base amino acid substitutions can affect the half life of the
molecule. In addition, engineered kinase dead forms of the LRRK2 protein (D1994S
mutant) have consistently been shown to be present at lower levels despite
comparable mRNA expression [238], suggesting the protein is less stable. LRRK2
G2019S has been demonstrated to have a half life comparable to the wild type
molecule in HEK293 LRRK2 over-expressing cells [201] suggesting the G2019S
mutation does not impact on LRRK2 stability. Furthermore, there is no evidence of
altered LRRK2 protein levels in G2019S LRRK2 PD brain tissue, although this
evaluation has been limited to semi quantitative immunohistochemical analysis of
amygdala, putamen, cerebellum and frontal cortex regions [331]. In support of these
immunohistochemical findings, the combined immunoprecipitation and Western
blot approach determined similar LRRK2 protein levels in control and G2019S
fibroblast lines. Although the level of LRRK2 expression was found to be two fold
lower in the G2019S LRRK2 over-expressing SHSY5Y cells relative to wild type LRRK2
229
levels, this difference is most likely due to clone to clone variation rather than the
processing of the mutant protein.
6.5 LRRK2 cellular localization
6.5.1 Wild type and G2019S LRRK2 cellular distribution
As well as having a broad tissue distribution, LRRK2 appears to have a broad
sub-cellular localization. Sub-cellular fractionation studies have detected LRRK2 in
mitochondria, lysosomal, ER and Golgi enriched compartments with greatest
proportion of cellular LRRK2 in the cytosolic and vesicular fraction as described for
ectopic LRRK2 in SHSY5Y cells [239], HEK293 cells and endogenous protein in rat
whole brain tissue homogenates [201, 251]. The proposed LRRK2 cellular
distribution is in line with the data obtained here where LRRK2 was detected in all
sub-cellular fractions with the greatest levels identified in the cytosolic and vesicle
enriched compartments of SHSY5Y over-expressing cells. Endogenous LRRK2 protein
in lymphoblasts had a comparable distribution pattern suggesting LRRK2 overexpression does not lead to gross changes in the localization of the protein. LRRK2
has been identified in synaptosomal isolations from mouse brain tissue with a
proposed role in vesicle handling [239, 254]. Identification of LRRK2 in the vesicle
enriched fraction in SHSY5Y cells provides a potential link between LRRK2
localization in the brain and synaptic function. Analysis by immunogold labelling and
electron microscopy of endogenous LRRK2 in rat basal ganglia identified LRRK2
immunoreactivity in mitochondria, autophagosomes, endosomal and Golgi transport
vesicles [251] supported by immunocytochemical analysis of over-expressed LRRK2
co-localization with ER, Golgi, lysosomal and mitochondrial markers in HEK293 and
SHSY5Y cells [218, 227, 239, 251]. However, the immunogold analysis was
performed without the corresponding LRRK2 knockout controls hence the specificity
of the antibody labelling is questionable. Immunohistochemical evaluation of wild
230
type LRRK2-V5 over-expression in SHSY5Y cells described in this report identified a
comparable pattern of diffuse antibody staining. Co-localization studies were not
carried out, however lysosomal, mitochondrial and ER localization was supported by
differential centrifugation analysis identifying 5 % of LRRK2 in cellular compartments
containing these organelle markers. In line with these observations, the direct
interaction of LRRK2 with multi vesicular bodies and mitochondria has been linked
to autophagic regulation of the endocytic lysosomal pathway in BAC-LRRK2
transfected HEK293 cells, as well as mediating mitochondrial function and
morphology in SHSY5Y cells and primary neurons [280, 332].
Pathological forms of LRRK2, including kinase domain (G2019S, I2020T) and
GTPase (R1441G, R1441C) mutants have previously been linked to the formation of
LRRK2 aggregates in HEK293 [215, 228, 298, 333] and SHSY5Y cells [224] overexpressing the protein. Aggregates in wild type LRRK2 over-expressing HEK293 and
SHSY5Y cells have also been noted albeit at lower levels. LRRK2 aggregation has
been linked to reduced cell viability [333] suggesting aberrant LRRK2 cellular
localization may be a feature of G2019S LRRK2 PD. Cellular fractionation
experiments identified similar LRRK2 localization for wild type and G2019S LRRK2 in
HEK293 and SHSY5Y cells [201, 239]) comparable to the data obtained in our study
comparing cellular distribution of wild type and G2019S LRRK2 by differential
centrifugation. In addition, although limited to analysis of the wild type protein, our
immunocytochemical evaluation of LRRK2 cellular localization did not reveal any
LRRK2 positive aggregates in SHSY5Y over-expressing cells. Aggregated LRRK2
structures have not yet been described in PD brains or cell models expressing
physiologically relevant levels of LRRK2 suggesting the observed aggregates could
occur as a result of LRRK2 over-expression above a certain threshold. The level of
231
LRRK2 over-expression relative to endogenous LRRK2 was not evaluated in any of
the studies reporting LRRK2 aggregates [215, 224, 228, 298, 333].
6.5.2 Wild type and G2019S LRRK2 mitochondrial localization
Previously, it has been estimated that 10 % of endogenous LRRK2 can be found
in the mitochondrial enriched fraction of rat brain mitochondria isolated by glycerol
density gradient [251]. LRRK2 mitochondrial localization has also been evaluated in
affinity purified mitochondrial preparations [280, 281] however estimations of
LRRK2 mitochondrial abundance were not carried out in these studies. In addition,
an accurate comparison of wild type and G2019S LRRK2 mitochondrial levels has not
yet been described. 10 % of total cellular LRRK2 was identified in the mitochondrial
enriched compartment in the sub-cellular fractionation experiments. Although
affinity purification and removal of lysosomal and ER contaminants confirmed that
LRRK2 was present in highly purified mitochondria, the level of mitochondrial LRRK2
may well be very low. As a large portion of cytochrome C was lost during the
mitochondrial affinity isolation, it is possible that the integrity of affinity purified
mitochondria was compromised and a fraction of the outer membrane bound LRRK2
[251] was lost. Nonetheless comparing levels of wild type and G2019S LRRK2 in
affinity purified preparations identified similar portions (1-2 %) of the total LRRK2
pool in the mitochondrial fraction further in support of comparable LRRK2
intracellular distribution for the wild type and G2019S protein.
ER and lysosomal contamination was not investigated in mitochondria isolated
from rat brain and may therefore contribute to the estimated 10 % of the LRRK2
mitochondrial pool [251]. However, mitochondrial LRRK2 levels in the rat brain may
also be different to the level of mitochondrial LRRK2 in the over-expressing SHSY5Y
cells. Cell specific LRRK2 mitochondrial localization is supported by the barely
detectable levels of endogenous LRRK2 in the mitochondrial enriched fractions
232
isolated from lymphoblasts. The presence of LRRK2 in highly purified mitochondria
from over-expressing SHSY5Y cells could relate to high LRRK2 levels, however these
were only 2-3 fold higher than endogenous expression in lymphoblasts, comparable
to the difference in LRRK2 cellular levels between wild type and G2019S overexpressing cells where similar levels of mitochondrial LRRK2 were identified. The
differences in mitochondrial LRRK2 levels between lymphoblasts and LRRK2 overexpressing SHSY5Y cells are probably not due to antibody sensitivity problems as the
combined immunoprecipitation and Western blot approach could detect LRRK2
immunoreactivity in diluted (1:10) fibroblast lysates containing 100 fold lower levels
of LRRK2 protein relative to the equivalent amount of undiluted lymphoblast input.
Rather the differences in mitochondrial LRRK2 abundance between the cell types
could indicate a cell specific functional regulation.
LRRK2 lacks an obvious mitochondrial localization sequence as has been
described for the PINK1 protein [46]. The mechanism of LRRK2 mitochondrial
targeting has not been investigated. Interactions with 14-3-3 proteins are recognized
mechanisms regulating the mitochondrial localization of certain proteins including
Bax [334]. LRRK2 is known to interact with 14-3-3 which is disrupted when LRRK2
kinase is inhibited [216]. However, inhibition of LRRK2 kinase activity disrupts 14-3-3
binding but did not influence mitochondrial LRRK2 levels suggesting 14-3-3 binding
was not regulating its mitochondrial localization.
6.5.3 LRRK2 localization in the mitochondria
25-50 % of the total LRRK2 pool is believed to be membrane associated
suggesting membrane association could play a role in the LRRK2 mitochondrial
interaction [252]. LRRK2 association with the outer mitochondrial membrane in
isolated rat brain mitochondria has previously been described [251]. In support of
this LRRK2 was shown to behave like a mitochondrial membrane bound protein as
233
determined by the partial release of LRRK2 from the mitochondria under alkaline
conditions. Hypotonic conditions did not appear to affect mitochondrial LRRK2
abundance further suggesting that mitochondrial LRRK2 was membrane associated
as the membrane unbound protein cytochrome C was released under these
conditions.
Sterically, an interaction between LRRK2 and the outer mitochondrial
membrane is most plausible. In addition, the full length 280 kDa molecule is too
large for import through transmembrane channels [335]. Our limited data supported
the LRRK2 association with the outer mitochondrial membrane however attempts to
accurately map the LRRK2 mitochondrial localization were hampered by a number of
technical issues. The shift of the inner membrane markers to the soluble fraction
while the outer membrane markers were present in the pellet suggested the outer
membrane was still intact when the inner membrane was already compromised,
observed for both crude and pure preparation. We also experienced problems with
the proteinase K digestion and the preferential removal of the outer membrane.
Loss of mitochondrial integrity, as determined by cytochrome C release during
mitochondrial affinity purifications and MEF isolations may have prematurely
exposed the inner membrane to digitonin and proteinase K. As both methods have
previously been successfully carried out on SHSY5Y cells [336, 337], additional
optimization of the technique may be required focusing on preserving the integrity
of mitochondrial isolations.
6.6 Wild type and G2019S LRRK2 kinase activity
LRRK2 has previously been demonstrated as an active kinase in in vitro
assays utilizing full length immunoprecipitated LRRK2 protein from HEK293 cells and
lymphoblasts as well as a truncated version of LRRK2 lacking the N terminus purified
from HEK293 cells (GST-LRRK2 1362-2527) [202], monitoring the incorporation of 32P
234
ATP into artificial substrates (LRRKtide) and LRRK2 interacting partners MBP [202]. In
the same assays the pathological G2019S LRRK2 mutant has consistently been
shown to have higher kinase activity when compared to the wild type protein [338]
while inhibition of LRRK2 kinase activity by LRRK2-IN1, CZC25146 and TAE-684 has
been shown to correlate with dephosphorylation of LRRK2 at Serine 910 and 935
[214, 215, 339]. LRRK2tide phosphorylation was detected in in vitro kinase assays of
LRRK2 immunoprecipitated from lymphoblasts and over-expressing SHSY5Y cells,
confirming these cell models expressed an active LRRK2 molecule. LRRK2 specific
activity was 3-5 times lower for LRRK2 immunoprecipitated from over-expressing
SHSY5Y cells relative to lymphoblasts suggesting LRRK2 is less active in this cell type.
Although the discrepancy may be due to cell specific differences, the C terminal V5
tag may have impacted on the kinase activity of the protein. In relation to this,
truncations of 7 C terminal residues in the LRRK2 WD40 domain have previously
been shown to ablate the kinase activity of LRRK2 suggesting the C terminus is
potentially involved in enzymatic regulation of the protein [202]. However, Serine
935 phosphorylation in wild type and G2019S LRRK2 over-expressing SHSY5Y cells
was sensitive to LRRK2 kinase inhibition as determined by Western blot analysis of
the immunoprecipitated protein further suggesting the tagged enzyme was active in
this cell type. In addition, the inhibitor doses that were required to reduce LRRK2
Serine phosphorylation by >90 % were comparable to those shown for enzymatically
active, N-terminally tagged LRRK2 immunoprecipitated from HEK293 cells [215]
suggesting LRRK2 kinase activities were similar in the two systems.
The kinase activity of G2019S LRRK2 immunoprecipitated from SHSY5Y overexpressing cells was 2 fold greater relative to the wild type protein in line with
previous observations [338]. The potency of LRRK2 IN1 was greater for G2019S
relative to wild type LRRK2 immunoprecipitated from SHSY5Y cells, implying G2019S
235
LRRK2 has greater affinity for ATP as LRRK2 IN1 is an ATP competitive inhibitor,
consistent with enzymatic hyperactivation associated with the pathogenic mutant.
The kinase activity of LRRK2 expressed in control and G2019S patient fibroblasts was
not evaluated due to the low level of LRRK2 expression in this cell type. LRRK2 kinase
activity in fibroblasts has not been reported in the literature, however the effects of
the G2019S mutation on LRRK2 kinase activity have been described for kidney and
immune cells lines [215] for endogenous and over-expressed LRRK2 suggesting the
increased kinase activity associated with the mutant protein is not dependent on the
cell type and G2019S fibroblasts are likely to have increased LRRK2 kinase activity
relative to control cells.
6.7 LRRK2 and mitochondrial function
Mitochondrial respiratory chain abnormalities are a common feature of
idiopathic, genetic PD and experimental models of the disease; reduced Complex I
activity has been described for idiopathic PD substantia nigra, platelets and skeletal
muscle [99, 340] suggesting respiratory function was compromised. The
parkinsonian phenotype induced by systemic MPTP injections in mice correlated
with a reduction in mitochondrial respiration [341, 342]. Various genetic causes of
PD have compromised mitochondrial function; Complex I defects have been noted in
fibroblasts harbouring PD associated PINK1 mutations [343] while reduced
mitochondrial respiration and mitochondrial membrane potential correlated with
the loss of PINK1 in mouse embryonic fibroblasts [344]. Parkin deficiency in
Drosophila melanogaster and PD patient derived parkin mutant fibroblasts were
associated with reduced oxygen consumption rates [345, 346]. α-synuclein overexpression in primary cells was shown to inhibit Complex I activity [123].
Mitochondrial membrane potential and ATP levels were reduced in SHSY5Y cells,
primary cortical cultures [280, 281] and fibroblasts [285] expressing G2019S LRRK2
236
consistent with abnormal respiratory chain function, suggesting respiratory chain
dysfunction may be a common feature in different forms of PD and potentially
involved in the disease pathogenesis.
Although variability was detected in ATP measurements for control and
G2019S fibroblasts described in this report, as a group G2019S cells had reduced ATP
levels, consistent with the previous observations [280, 281, 285].
In addition,
G2019S LRRK2 expression correlated with a 20 % reduction in mitochondrial
membrane potential in SHSY5Y and fibroblast cell models. The reduction in
mitochondrial membrane potential in SHSY5Y cells was comparable to what has
previously been described [280, 281], while fibroblast showed a more modest defect
in comparison to the 60 % decrease determined in the previous report [285]. Of
note, Mortiboys et al, determined an 85 % reduction in the mitochondrial
membrane potential in control fibroblasts treated with parkin siRNA [347], a
decrease much greater than those generally reported for viable cells. Reductions of
15-20 % in the mitochondrial membrane potential have previously been observed
for fibroblasts derived from patients with PINK1 mutations and PINK1 deficient
mouse embryonic fibroblasts [172, 348] consistent with the magnitude of changes
reported for our G2019S expressing fibroblasts. In addition, fibroblasts have
previously been shown to adjust to gross mitochondrial defect such as complex I
enzymatic deficiency by reversing the mitochondrial ATPase [310]. We have
modelled this in control fibroblasts exposed to rotenone and demonstrated ATPase
reversal by the response of the rotenone treated cells to oligomycin which led to the
dissipation of the mitochondrial membrane potential. Hence a 60 and 85 % decrease
in mitochondrial membrane potential is physiologically unlikely as the fibroblasts
would most likely attempt to adjust to such a massive perturbation in the
mitochondrial proton gradient. The discrepancy in the mitochondrial membrane
237
potential measurements is therefore likely due to an over-estimation of the
magnitude of the defect. Nonetheless a common trend was observed among all the
LRRK2 studies detecting a reduction in the mitochondrial membrane potential and
ATP levels associated with G2019S LRRK2.
Oxygen consumption measurements detected a 1.5-2 fold increase in
cellular respiration rates in G2019S expressing fibroblasts and SHSY5Y cells. The
increase in cellular respiration in G2019S expressing cells opposed the respiratory
deficiencies reported for PINK1 deficient mouse embryonic fibroblasts [348], MPTP
and rotenone treated mice [342] and parkin deficient Drosophila melanogaster
[345].
The reduction in basal ATP levels and mitochondrial membrane potential
have previously been assigned to a mitochondrial defect associated with G2019S
LRRK2 expression consistent with the respiratory chain abnormalities linked with the
genetic and experimental PD models [280, 281, 285]. Although the decrease in ATP
levels and mitochondrial membrane potential detected in G2019S expressing
fibroblasts and SHSY5Y cells mirrored the characteristics of bioenergetically
compromised cells, increased oxygen utilization argued against a respiratory chain
defect. Measurements of chemically uncoupled respiration revealed comparable
cellular respiratory capacities for control and G2019S fibroblasts, further arguing
against a mitochondrial dysfunction.
Mitochondrial respiration of fibroblasts cultured under substrate limiting
conditions was comparable to those observed in G2019S expressing cells suggesting
mutant LRRK2 expressing cells could suffer from deficient glycolytic input, as has
previously been shown for PINK1 deficient neuroblastoma cells and mouse primary
dopaminergic neurons [172]. However, mitochondrial membrane potential was
increased in fibroblasts cultured on galactose consistent with an increased flux
238
though the mitochondrial respiratory chain, different to the effect on the
mitochondrial membrane potential observed in G2019S expressing fibroblasts and
SHSY5Y cells. In addition, the respiratory chain defect described in PINK1 and parkin
deficient cells has been associated with mitochondrial ATPase reversal in primary
cortical neurons and SHSY5Y cells [172] as determined by the dissipation of the
mitochondrial membrane potential in the presence of oligomycin. Oligomycin
treatment produced comparable effects on the mitochondrial membrane potential
in control and G2019S fibroblasts consistent with a normal functioning ATPase. As
there was no evidence for limitation in glycolytic substrate supply or mitochondrial
ATPase reversal in G2019S LRRK2 expressing fibroblasts and SHSY5Y cells these
systems are likely to have a different underlying cause to the mitochondrial
bioenergetic changes associated with mutant LRRK2. In fact, the combination of
reduced mitochondrial membrane potential and enhanced respiration supported
the presence of a mitochondrial proton leak [349].
Bioenergetically, a mitochondrial proton leak has previously been
characterized by oxygen utilization measurements in the presence of oligomycin; a
respiration rate was detected in oligomycin treated thymocytes over-expressing
UCP2 [350] absent in control cells. Oligomycin inhibited respiration rates have also
been shown to contribute to skeletal contractions in mice [351]. In support of a
mitochondrial proton leak, oxygen utilization was taking place in G2019S expressing
fibroblasts in the presence of oligomycin, absent in control cells.
6.8 LRRK2 and mitochondrial permeability
The decrease in mitochondrial membrane potential combined with
increased oxygen utilization rates associated with G2019S expression implied an
increased permeability of the mitochondrial inner membrane to protons. This may
involve opening of the mitochondrial permeability pore [352], the presence of
239
proteins on mitochondrial forming a pore [353] or the up-regulation and activation
of uncoupling proteins [354].
Opening of the mitochondrial permeability pore is associated with increased
reactive oxygen species (ROS), increased mitochondrial calcium levels and apoptosis
[352]. Respiratory chain Complex I defects, a prominent feature in PD patient brains
and skeletal muscle [99] as well as PINK1 deficient fibroblasts [343], have correlated
with increased oxidative stress, apoptosis and mitochondrial permeability pore
opening [348]. Wild type and mutant LRRK2 over-expression in SHSY5Y cells has
previously been shown to promote PTP opening through the interaction between
adenine nucleotide translocator (ANT) and porin [286]. In addition, SHSY5Y cells
over-expressing the G2019S LRRK2 protein showed increased cytochrome C release,
the activation of the caspase cascade and increased cell death [227]. However, there
was no evidence of cell death in our G2019S expressing fibroblasts or SHSY5Y cells.
In agreement with this thesis, oxidative stress levels were comparable between wild
type and G2019S over-expressing SHSY5Y cells in one report [280] but in another
study greater for G2019S expressing SHSY5Y cells [281]. Oxidative stress was also
reported in G2019S expressing fibroblasts relative to control cells [285] however the
rate of ROS generation in G2019S fibroblasts described in this report were lower
relative to control cells. Hence the effects of G2019S expression on ROS generation
are not consistent and may vary depending on cell type and assay conditions. The
absence of cell death and increased ROS levels suggested the mitochondrial
permeability transition pore was unlikely to cause the mitochondrial uncoupling in
G2019S expressing fibroblasts and SHSY5Y cells.
Regarding any direct interaction of LRRK2 with mitochondria leading to
increased proton permeability of the inner membrane, we confirmed that a small
proportion of LRRK2 was present in highly purified mitochondria. Further
240
mitochondrial sub-fractionation suggested LRRK2 associated with mitochondrial
membranes, however, we were not able to restrict this to either the outer or inner
membranes. The mitochondrial uncoupling in G2019S SHSY5Y cells was not likely to
be an artefact of over-expressed LRRK2 localizing to the mitochondria as uncoupling
was also observed in fibroblasts expressing endogenous G2019S LRRK2 and absent
in SHSY5Y cells over-expressing the wild type protein.
The mitochondrial uncoupling associated with G2019S mutant fibroblasts
and SHSY5Y cells closely resemble what has previously been described for cells and
tissues with increased uncoupling protein expression. Increased oxygen
consumption rates, reduced mitochondrial membrane potential and ROS production
have been reported for mitochondrial isolations from mouse brains, livers and
kidneys of UCP2 transgenic mice [355] as well as CHO cells over expressing UCP1
[356]. Reduced membrane potential has been demonstrated for mitochondria in
UCP4 over-expressing cells [357] and UCP4/5 reconstituted in liposomes [358]. In
line with these observations, we observed a reduced mitochondrial membrane
potential in SHSY5Y cells over-expressing the UCP2-GFP protein.
6.9 LRRK2 regulation of mitochondrial uncoupling proteins
6.9.1.1 LRRK2 regulation of UCP mRNA expression
Mitochondrial uncoupling has been demonstrated to be regulated by
specific UCP isoforms in different cell types; UCP1 is the dominant isoform in brown
fat reported to regulate thermogenesis [359], while the UCP2 isoform is found in
pancreatic β-cells and mediates insulin secretion through ATP mediated opening of
potassium channels [360]. Expression of four UCP isoforms (UCP2-5) was detected in
control fibroblasts and SHSY5Y cells, suggesting these proteins may have a functional
role in these cell lines. UCP2, 4 and 5 mRNA expression has been reported in
fibroblasts, although UCP3 mRNA expression was not investigated [361]. UCP2 has
241
previously been identified as the abundant UCP isoform in the rodent brain,
expression of UCP4 and 5 was also described consistent with the pattern of UCP
expression detected in SHSY5Y cells [362].
The LRRK2 G2019S mutation correlated with an increase in UCP2-5 mRNA in
fibroblasts and increased UCP4 expression in SHSY5Y cells. The UCP mRNA changes
observed in G2019S expressing fibroblasts and SHSY5Y cells suggested the LRRK2
associated mitochondrial uncoupling was regulated at the transcriptional level.
Increased UCP4 mRNA expression has previously been demonstrated in SHSY5Y cells
[363] and brain in experimental stroke models [364] while UCP2 expressional
changes in response to PGC1α over-expression have been described in myotubes,
confirming that these isoforms can be transcriptionally regulated. Of note, increased
UCP4 levels in G2019S LRRK2 SHSY5Y cells suggested the artificial UCP2 overexpressing system used in this study may not be an accurate model of uncoupling
for this cell type.
The fold changes in UCP4 mRNA expression were comparable to 5 fold
changes reported for UCP1 expression in brown fat comparing tissue isolated from
warm acclimated mice to those exposed to cold temperatures [365]. A similar 2-3
fold induction in UCP3 mRNA has been described in adipocytes and skeletal muscle
associated with reduced body weight in mice suggesting small changes in UCP mRNA
levels are sufficient to induce phenotypic changes [366].
6.9.1.2 LRRK2 regulation of UCP protein
Changes in UCP2 and 3 mRNA have previously been shown to affect UCP
protein levels in pancreatic β cells [367]. While ectopic UCP2 expression in the UCP2GFP SHSY5Y model was confirmed by confocal microscopy and Western blot,
Western blot analysis detected additional higher and lower molecular weight
immunoreactive species making it difficult to identify and quantitively evaluate
242
endogenous UCP2 protein in wild type and G2019S fibroblasts and SHSY5Y cells.
Cross reactive bands of 28 and 50 kDa have previously been described for UCP2
over-expressing insulinoma cells [368] and in a variety of mouse and rodent tissues
[368] suggesting antibody specificity may not be reliable enough to distinguish
between true endogenous species.
To determine the potential contribution of UCP protein to the mitochondrial
uncoupling observed in G2019S fibroblasts and SHSY5Y cells, UCP activity was
inhibited by genipin. Although genipin is often described as a general UCP inhibitor,
the characterization of the drug has been limited to in vitro studies involving UCP2
protein [315] and its selectivity profile for the other UCP family members has not
been fully elucidated. Studies of pancreatic islet cells have previously shown genipin
to increase the mitochondrial membrane potential, ATP levels and promote insulin
secretion, a function governed by UCP3 [315] while inhibition of mitochondrial
proton leak in leukemic MX2 cells correlated with lower ROS levels and reduced
drug resistance of the cancer cells [369] linking UCP2 protein activity and inhibition
to respiratory chain function.
Consistent with UCP mRNA expression in control SHSY5Y cells and
fibroblasts, a mild 10 % increase in mitochondrial membrane potential was observed
upon genipin treatment suggesting a role for UCP proteins in the regulation of
respiratory chain function. In addition, restoration of the mitochondrial membrane
potential in UCP2-GFP cells with genipin suggested the drug could counteract the
phenotype of UCP2 protein over-expression. The specificity of genipin for the UCP2
isoform was questioned by the effects of the compound on the G2019S LRRK2 overexpressing SHSY5Y cells, where the UCP4 mRNA expression was preferentially
induced. As UCP3 and 5 were found to be expressed in SHSY5Y cells and UCP2 mRNA
was expressed at greater levels relative to UCP4 mRNA, the observed effect could be
243
due to genipin inhibition of all the UCPs in SHSY5Y cells. UCP2 mRNA levels were
comparable for control, wild type and G2019S over-expressing clones however, we
cannot rule out the possibility of increased UCP2 protein in SHSY5Y cells expressing
G2019S LRRK2. Although it was not possible to specifically inhibit UCP2 and UCP4 in
fibroblasts and SHSY5Y cells respectively, restoration of the mitochondrial
membrane potential in G2019S expressing cells upon genipin treatment supported
the role of UCP protein in the mitochondrial uncoupling associated with mutant
LRRK2.
Of note, effective genipin concentrations reported here were considerably
lower with respect to ranges described in the literature. Genipin at 5 µM was
required to restore the UCP dependent contractile response in cardiac myocytes
[370]. Genipin at 10 µM was determined as the effective dose for toxicity in
leukemic cells [369] while 200 µM genipin induced toxicity in hepatoma cells [371].
Although toxicity was not apparent in our short term genipin incubations for SHSY5Y
cells, the mitochondrial membrane potential in control fibroblasts was reduced by
20 % in prolonged treatments (30 min) suggesting this cell type is more sensitive to
the drug. The cell toxicity associated with genipin treatment is likely due to the off
target effects associated with the drug. In support of this, there is no evidence to
suggest siRNA knockdown of UCPs 2 and 4 lead to reduced cell viability [363, 372].
Effective genipin dose requirements for UCP inhibition in different cell types could
potentially correlate with the levels of expression of the different UCP isoforms.
6.9.1.3 LRRK2 regulation of PGC1α expression
The difference in the relative mRNA abundance between the UCP isoforms
in SHSY5Y cells and fibroblasts as well as induction of specific UCPs (2 and 4) in these
cells may reflect cell specific regulatory roles such as those described for adipose
tissue and the regulation of thermogenesis by UCP1 and pancreatic β cells in the
244
modulation of insulin secretion by UCP2 [359, 360]. Tissue specific UCP4 and 5
regulation has also been described for dopaminergic neurons in the SNpc, required
for the regulation of pacemaking activity [168]. Therefore LRRK2 regulation of
different UCP isoforms in fibroblasts and SHSY5Y cells may be required for cell
specific mitochondrial uncoupling mediated functions.
Cell specific UCP mRNA expression has previously been linked to PGC1α
mediated transcriptional regulation. Stimulation of PGC1α expression was described
for brown adipocytes undergoing differentiation and developing muscle, processes
requiring induction of UCP1 and UCP2 transcription [373, 374]. In these studies
UCP1-3 expression was detected in both cell types however only UCP1 or 2 mRNA
was preferentially stimulated in the two studies [373, 374]. Increased UCP4 mRNA
levels correlated with a two fold increase in PGC1α expression in G2019S LRRK2
over-expressing cells suggesting PGC1α may play a role in the specific regulation of
UCP4 expression.
PGC1α mediated UCP2 regulation has previously correlated with an increase
in catalase and SOD2 mRNA in endothelial cells undergoing oxidative stress [141]. In
addition, increased mitochondrial densities in white adipose tissue [375] and muscle
cells [376, 377] correlated with PGC1α mediated UCP1, 2, TFAM and Nrf1, 2
expression in line with stimulation of mitochondrial biogenesis suggesting the
increase in UCP2 and UCP4 expression could be a part of a biosynthetic or
antioxidant gene response. The relative expression of antioxidant enzymes SOD2
and catalase were comparable between control, wild type and G2019S expressing
cells suggesting the increased UCP4 expression associated with the G2019S
mutation was not a part of a global antioxidant response. Furthermore comparable
abundance of mitochondrial proteins, TFAM and mitochondrial structures in control
and mutant fibroblasts and SHSY5Y cells implied rates of mitochondrial biogenesis
245
were not perturbed. The absence of an oxidative stress response and changes to
mitochondrial abundance implies the induction of UCP expression in the presence of
G2019S is unlikely a stress response and selectively targets the UCP gene.
The selectivity for UCP4 induction could potentially be accounted for by
PGC1α interacting molecules in the relevant cell types. PGC1α mediated UCP
response in skeletal and cardiac muscle as well as brown adipose tissue has
previously been shown to be dependent on the isoform of its interacting gene
regulator PPAR. UCP3 expression in skeletal muscle is more responsive to the PPARδ
than PPARα isoform [378, 379] and UCP3 mRNA induction in cardiac muscle requires
PPARα [380]. While induction of UCP2 expression in skeletal muscle and heart is
responsive to PPAR related stimuli in a similar manner to UCP3, additional
regulatory mechanisms exist [381]. PGC1α in muscle cells requires co-activators
MyoD and p300 for UCP2 gene stimulation [382]. In this study, the direct regulation
of the UCP promoter by PGC1α was not investigated and warrants further analysis to
determine whether a specific PPAR isoform and additional regulators are involved.
The selective changes in UCP4 mRNA further emphasise the importance of UCP
regulated mitochondrial uncoupling in G2019S LRRK2 expressing fibroblasts and
SHSY5Y cells.
6.9.1.4 LRRK2 regulation of HDAC5
Based on the sub-cellular localization data for both endogenous and overexpressed LRRK2, a large portion of the full length protein pool appeared to exist in
the cytosolic compartment. The multiple structural domains of LRRK2 implicate a
scaffold function in signalling cascades and the reported association with 14-3-3
[215, 216, 383] suggested LRRK2 may potentially be involved in compartmental
targeting of 14-3-3 interacting molecules while 14-3-3 could in turn regulate LRRK2
function. LRRK2 mediated regulation of nuclear to cytosolic shuttling has been
246
described for the nuclear factor of activated T cells (NFAT) in T cells where LRRK2
was identified as a crucial component of the NRON complex maintaining NFAT in its
inactive state in the cytosol [271]. Loss of LRRK2 was linked to the translocation of
NFAT to the nucleus and hyperactivation of transcription. HDAC5 mediated PGC1α
transcription has previously been shown to also involve nuclear translocation in
response to protein kinase A (PKA) and protein kinase C (PKC) [317]. Hypothesising a
mechanistic conservation, LRRK2 association with HDAC5 was investigated to
determine what effect LRRK2 kinase activity has on the cellular localization and
integrity of this complex and how the potential LRRK2 HDAC5 interaction could be
involved in PGC1α transcription and UCP expression.
In an attempt to determine what effect LRRK2 G2019S has on HDAC5
nuclear translocation, HDAC5 subcellular localization was investigated. Under basal
conditions HDAC5 has been shown to localize to the nucleus and cytoplasm with the
relative nuclear to cytosolic ratios depending on the cell system in question [317,
384]. HEK293 and HeLa cell lines have nuclear HDAC5 abundance while the
localization of HDAC5 in H9C2 rat heart cell line was shown to vary depending on the
differentiation status [385]. In addition, PMA treatment of cells in these studies has
been shown to mobilize the HDAC5 pool and result in the complete redistribution of
HDAC5 to the cytoplasm. HDAC5 nuclear export in response to PMA has been
demonstrated
in
Cos7
cells
[317,
384].
The
antibodies
used
for
immunocytochemistry have been limited to HDAC5 over-expressing systems with no
reports of endogenous HDAC5 localization in SHSY5Y cells yet described [317, 384].
We detected endogenous HDAC5 in the nucleus and cytosol of SHSY5Y cells by
immunocytochemistry, however PMA treatment of wild type cells did not grossly
affect the nuclear HDAC5 pool suggesting antibody specificity may be an issue as the
same PMA conditions have consistently been shown to lead to HDAC5 nuclear
247
export in SHSY5Y cells [317]. As we did not have a suitable HDAC5 knockout system
to validate the HDAC5 antibody for immunohistochemistry, analysis of HDAC5
cellular distribution in wild type and G2019S LRRK2 over-expressing SHSY5Y cells was
not carried out.
Immunoprecipitation of the HDAC5 protein has previously been carried out
on Flag tagged, over-expressed species in COS7 cells [386] and GFP tagged HDAC5 in
SHSY5Y cells [387], detecting 120 kDa immunoreactive bands corresponding to the
full length tagged molecule. Although the relative molecular mass of the HDAC5
immunoreactive band detected in HDAC5 immunoprecipatates matched the
molecular weight of the predicted and previously described full length HDAC5, an
additional lower molecular weight species was also detected which has not
previously been reported for IPs of the HDAC5 protein. To determine whether the
80 kDa species detected here is an artefact associated with the HDAC5 antibody,
additional characterization of the antibody should be carried out on HDAC5
knockout tissue. Nevertheless LRRK2 was detected in HDAC5 immunoprecipitated
from wild type and G2019S LRRK2 over-expressing SHSY5Y cells and wild type but
not LRRK2 knockout mouse brain cortex.
Immunoprecipitation of LRRK2 has previously identified NFAT in T cells
[271], Ago2 in the aged fly brain [273], 14-3-3 in HEK293 cells and lymphoblasts
[216], DLP1 in SHSY5Y cells and primary cortical neurons [280, 281] as potential
LRRK2 interacting molecules. LRRK2 interaction with DLP1 was enhanced for the
G2019S LRRK2 mutant in SHSY5Y cells [280], the human Ago2 protein pulled down
more of the G2019S LRRK2 than the wild type [273], LRRK2 kinase inhibition has
previously been shown to correlate with the amount of bound 14-3-3 [215, 216]
suggesting LRRK2 kinase activity may play a role in complex formation. However,
although 14-3-3 binding to LRRK2 was affected by LRRK2 kinase inhibition, the
248
G2019S mutation did not impact on the amount of 14-3-3 associated with LRRK2
[215] suggesting additional regulatory factors may be involved in this interaction and
the regulation of downstream signalling cascades. Association of NFAT with the
NRON complex required the LRRK2 protein but the affinity of the interaction did not
seem to be affected by the G2019S mutation [271] suggesting the integrity of
complexes involving LRRK2 may not always depend on LRRK2 kinase status. The
intensity of the LRRK2 immunoreactive species detected in the HDAC5 pulldowns
was consistently stronger in G2019S LRRK2 over-expressing SHSY5Y lysates relative
to wild type LRRK2. The efficiency of HDAC5 pulldown was similar between control,
wild type and G2019S LRRK2 over-expressing suggesting that the G2019S LRRK2
protein was more associated with HDAC5, similar to the effects described for
G2019S LRRK2 association with Ago2 and DLP1 proteins. Although not investigated
in the studies involving Ago2 and DLP1, LRRK2 kinase inhibition did not seem to
affect the integrity of the HDAC5 interaction, in contrast to what has previously been
shown for 14-3-3 suggesting Serine 910 and 935 phosphorylation is not required for
HDAC5 binding to LRRK2.
We reasoned that any potential interaction with LRRK2 could be enhanced
by enriching the cytosolic pool of HDAC5. PMA treatment of wild type and G2019S
LRRK2 over-expressing SHSY5Y cells failed to detectibly enhance cytosolic HDAC5 as
determined by Western blot analysis of IP input lysates, further questioning the
antibody specificity. PMA treatment of wild type and G2019S LRRK2 over-expressing
cells resulted in the loss of LRRK2 immunoreactivity in the HDAC IPs suggesting a
dissociation of the HDAC5 LRRK2 complex. LRRK2 interaction with 14-3-3 has
previously been shown to be disrupted upon de-phosphorylation of Serine residues
[215, 216] in a PKA dependent mechanism [383]. As PMA activates a number of
intracellular kinases and phosphatases including PKA [388], it is possible that the
249
intrinsic phosphorylation status of LRKR2 could be affected in the presence of the
drug. Although LRRK2 phosphorylation at Serine 935 was not affected by PMA
treatment, dephosphorylation of other serine residues [202, 383] cannot be ruled
out. Identification of LRRK2 specific phosphatases and determination of their
mechanism of action may provide a better understanding of LRRK2 interacting
partners and the associated complex formation.
LRRK2 association with HDAC5 suggested a potential role for LRRK2 in
HDAC5 mediated transcriptional regulation of PGC1α and UCP4 in SHSY5Y cells.
Increased association between G2019S LRRK2 and HDAC5 could potentially account
for increased PGC1α and UCP4 mRNA in SHSY5Y cells. Although LRRK2 kinase
inhibition did not influence the HDAC5 association, UCP4 expression was reduced in
G2019S expressing SHSY5Y cells suggesting UCP4 transcription was altered while the
LRRK2 HDAC5 association was still intact. An additional level of HDAC5 regulation
could involve post translational modification [384] as has been previously described
for NFAT and DLP1 proteins [389, 390]. Although DLP1 phosphorylation was
comparable for HeLa cells transfected with wild type and G2019S LRRK2, specific
phospho residues were not examined and site specific modification by G2019S
LRRK2 cannot be ruled out. Furthermore, the contribution of LRRK2 GTPase activity
could also contribute to the regulation of the HDAC5 interaction.
6.9.1.5 LRRK2 GTPase mutations and mitochondrial function
LRRK2 GTPase mutations have been described in North American, Basque,
English and Italian cohorts with the clinical features and pathology resembling that
of idiopathic and G2019S LRRK2 linked PD [28, 29, 391]. Similar to the LRRK2 G2019S
mutation, transgenic mice expressing R1441C do not have a neurodegenerative or
kidney phenotype [233, 240]. Studies investigating the interplay between the two
enzymatic domains are inconsistent with some groups describing GTPase LRRK2
250
mutations to decrease LRRK2 kinase activity in HEK293 and neuro2A cells [203, 204]
while studies carried out in HEK293 cells in two other studies suggest GTPase
mutations have no effect on the kinase activity of the protein [201, 202]. GTPase and
kinase mutations have been shown to increase LRRK2 affinity for GTP in HEK293
cells [203], however kinase deficient LRRK2 has normal GTPase activity [204]. The
effect of LRRK2 kinase inhibition on GTPase function has not besn studied. In
addition the Serine 935 phosphorylation status of LRRK2 is reduced in GTPase but
not LRRK2 G2019S mutants [215]. The enzymatic crosstalk between the kinase and
GTPase domains as well as the pathological and clinical similarities associated with
GTPase and kinse LRRK2 mutant proteins suggest the existence of converging
pathways in the disease pathogenesis.
Although over-expression of mutant kinase and GTPase LRRK2 proteins in
SHSY5Y, HEK293 cells and rodent primary neurons is associated with increased
toxicity [222] [203, 227, 260], functional studies have often struggled to identify a
common converging phenotype; reduced complexity of dopaminergic midbrain
neurons was described for G2019S but not R1441C transgenic mice [233], GTPase
mutations enhanced filament formation and promoted association with
microtubules in HEK293 cells, an effect not observed for LRRK2 G2019S [333], kinase
but not GTPase mutations increased Tau phosphorylation [263]. These studies
implicated the existence of pathways which are kinase independent but require
LRRK2 GTPase activity and visa versa, questioning their contribution to the common
PD pathology observed for the kinase and GTPase mutations.
Mitochondrial fragmentation and the reduced mitochondrial membrane
potential described [280, 281] was extended to the R1441C GTPase mutation
suggesting mitochondrial function is affected by kinase and GTPase mutant LRRK2
proteins. In addition, GTPase LRRK2 mutants showed increased affinity for the DLP1
251
protein at the mitochondrial membrane of SHSY5Y cells [280] suggesting the GTPase
mutation can affect the affinity of complex formation in a similar manner to G2019S
LRRK2. Mitochondrial uncoupling associated with G2019S LRRK2 and the increase in
UCP4 mRNA expression was described as a LRRK2 kinase dependent mechanism
based on in vitro kinase activity assays characterizing G2019S LRRK2 as a hyperactive
kinase and the effects of LRRK2 kinase inhibitors on UCP4 levels in SHSY5Y cells.
Whether LRRK2 GTPase activity contributes to the regulation of UCP expression was
not investigated. However, the proposed mechanism of HDAC5 contribution to the
regulation of PGC1α and UCP4 mRNA levels by LRRK2 demonstrated increased
affinity of G2019S LRRK2 for HDAC5 suggesting a structural modulation of the
complex, comparable to the effects of G2019S and R1441C on the DLP1 interaction.
LRRK2 kinase inhibition did not seem to disrupt the LRRK2 HDAC5 association. As
LRRK2 kinase deficient mutants retain GTPase activity [204] GTPase function may be
sufficient to maintain the interaction. As the PD associated GTPase R1441C mutation
has been shown to contribute to mitochondrial abnormalities in cell models of PD,
the effects of GTPase mutants on the regulation of the UCP4 transcription, PGC1α
expression and HDAC5 association warrants further investigation.
6.9.2 LRRK2 regulation of DJ1
Mitochondrial uncoupling is believed to mediate the pacemaker activity of
SNpc dopaminergic neurons while the loss of the PD associated DJ1 protein results
in the dysregulation of pacemaking in mouse SNpc [168]. The proposed mechanism
of regulation of the pacemaker activity was shown to involve UCP4 and 5. The level
of these isoforms correlated with oscillations in the mitochondrial membrane
potential. UCP4 and 5 levels were reduced in DJ1 knockout mice while the oscillatory
activity of pacemaking was inhibited in these animals. As LRRK2 was shown to
increase UCP2 and 4 in fibroblasts and SHSY5Y cells, it was possible to speculate that
252
LRRK2 regulated UCP mRNA through a genetic interaction with DJ1 potentially
contributing to the regulation of pacemaking in SNpc. In addition to UCP mRNA
regulation, DJ1 has also been shown to be involved in PGC1α mediated transcription
by binding to upstream regulatory elements of PGC1α target genes [145] in line with
PGC1α involvement in UCP4 regulation in G2019S LRRK2 SHSY5Y cells. Similar to the
UCP transcriptional changes in G2019S expressing SHSY5Y cells, UCP gene regulation
by DJ1 appeared specific as SOD2 and catalase expression was unperturbed in the
SNpc of DJ1 knockout mice, [168], further implicating LRRK2, DJ1 and PGC1α in a
common regulatory pathway involving UCP transcription.
To investigate whether the regulation of UCP4 expression in SHSY5Y cells
was mediated by DJ1, DJ1 knockdown cell lines were studied. Genetic knockdown of
DJ1 in SHSY5Y cells did not affect UCP4 or 5 expression in contrast to what has
previously been described for SNpc of DJ1 knockout animals [168] suggesting LRRK2
regulation of UCP4 in SHSY5Y cells is unlikely to be DJ1 dependent. As the functional
role of UCP4 and 5 was assigned to the regulation of the pacemaker activity in SNpc
neurons, the absence of pacemaking in SHSY5Y cells could eliminate the need for
this regulatory process. In support of this UCP4 and 5 expression in the neighbouring
ventral tegmental neurons lacking intrinsic pacemaker function was not affected by
the loss of DJ1 in mice [168].
UCP3 levels were reduced by 50 % in both lines of DJ1 knockdown SHSY5Y
cells suggesting a potential regulatory role of DJ1 for this UCP isoform in the
neuroblastoma cells. UCP3 specific regulation implies a specific functional response
as has previously been described for UCP3 expression in cardiac muscle cells
mediating contractile responses and adipose tissue regulating body mass [359, 392].
As the G2019S mutation was associated with changes in UCP4 expression, we did
not pursue with UCP transcriptional rescue experiments involving DJ1 and LRRK2 in
253
the SHSY5Y cells. However, a converging pathway of UCP mRNA regulation by LRRK2
and DJ1 in SNpc neurons should not be ruled out and warrant further investigation.
6.9.3 LRRK2 regulation of NFκB
As well as transcriptional regulation of UCP mRNA through the PGC1α
pathway, UCP mRNA expression has been demonstrated to be regulated through
activation of the NFκB pathway. NFκB mediated UCP4 regulation has been shown in
SHSY5Y cells in response to oxidative stress [357] suggesting UCP levels can be
mediated by NFκB. In support of this an NFκB response element has been described
on the UCP gene [357]. NFκB activation of UCP4 transcription in SHSY5Y cells occurs
in parallel with SOD2 [393, 394] as a mechanism to protect against increased levels
of cellular ROS [357, 393, 394]. LRRK2 has been demonstrated to regulate NFκB
mediated transcription; over-expression of LRRK2 in HEK293 cells increased NFκB
dependent transcription [243], shRNA knockdown of LRRK2 reduced NFκB mediated
transcription [395] while wild type, and G2019S LRRK2 equally enhanced NFκB
mediated transcription [395]. Although NFκB regulation of UCP4 was not
investigated in SHSY5Y cells, SOD2 levels between control, wild type and G2019S
LRRK2 over-expressing cells were similar suggesting NFκB mediated regulation was
not involved. However UCP4 gene regulation by NFκB independent of SOD2 cannot
be ruled out.
6.9.4 UCP4 knockdown and LRRK2 regulation of Argonate 2
In an attempt to reduce UCP4 levels in SHSY5Y, UCP4 siRNA knockdown was
carried out. UCP4 knockdown has previously been linked to perturbations in calcium
homeostasis in SHSY5Y cells [363] suggesting reduced UCP4 levels can affect the
coupling state of mitochondria, in line with what was reported in SHSY5Y cells
treated with LRRK2 siRNA. A similar efficiency of UCP4 siRNA knockdown (60%) in
SHSY5Y cells was described here correlating with an increase in TMRM intensity
254
suggesting a potential link between UCP4 mRNA expression and the mitochondrial
membrane potential. UCP4 protein levels were not evaluated in either study due to
the lack of effective antibodies. Although the stability of the UCP4 protein has not
been investigated, the UCP2 and 3 homologues have been reported to have a short
half life of 1-4 hours [354, 396, 397] suggesting short term silencing protocols reduce
UCP protein and affect cellular functions mediated by UCP levels. In support of this,
a 55 % reduction in UCP2 mRNA in islet cells correlated with a 25 % decrease in
mitochondrial membrane potential, 30 % reduction in respiration rates and a 20 %
increase in ATP levels [372] suggesting small changes in UCP mRNA expression can
have pronounced effects on respiratory chain function.
The G2019S mutation has previously been shown to impact on LRRK2
regulation of the Ago2 containing RNA-induced silencing complex (RISC) mediating
microRNA regulation of mRNA translation [273]. As this complex is also required for
the siRNA interaction with its target genes it is possible to speculate that the
efficiency of knockdown in the LRRK2 over-expressing cells is affected by LRRK2
involvement in the transcriptional regulation pathway. Similar LRRK2 siRNA
conditions resulted in a knockdown of lower efficiency in wild type (40 %) and
G2019S (25 %) LRRK2 over-expressing SHSY5Y cells. siRNA knockdown efficiencies in
G2019S LRRK2 cell culture models have not previously been described and warrant
further investigation. LRRK2 association with the Ago2/RISC complex and the
proposed effects of G2019S on microRNA regulation could also potentially be
involved in microRNA mediated regulation of UCP mRNA levels. In support of this,
microRNA regulation of UCP2 mRNA expression has been described in cardiac and
skeletal muscle required for muscle development [398]. However, the influence of
this pathway on UCP expression in fibroblasts and SHSY5Y cells was not investigated.
255
6.9.5 LRRK2 regulation of UCP phosphorylation
As well as regulation through transcription, UCP activity can be regulated at
the protein level; increased mitochondrial uncoupling has correlated with increased
phosphorylation of UCP1 in brown adipose tissues of cold acclimatised animals [159]
while glutathionylation of cysteine residues in UCP2 and 3 has been shown to
regulate mitochondrial respiration in thymocytes [399]. Although phosphorylation
mediated regulation has been described for the UCP1 isoform, the serine residue is
conserved among the family members [159] suggesting other UCP isoforms could
potentially be regulated in a similar fashion. The kinase associated with UCP
phosphorylation has not yet been described. LRRK2 has previously been shown to
phosphorylate tau and 4E-BP functionally linked to mediating tubulin polymerization
and induction of autophagy [231, 261]. Identification of LRRK2 in the mitochondrial
enriched fraction of SHSY5Y cells suggested the protein could potentially mediate
UCP activity by post translational regulation. To directly test LRRK2 phosphorylation
of UCP2 in vitro kinase assays were carried out using recombinant proteins but failed
to produce any detectable UCP2 phosphorylation. However, as UCP functional
activity in vitro is enhanced by protein incorporation into liposomes [358, 400], it is
also possible that the crude assay conditions presented in this study require
modification.
6.10 LRRK2 regulation of mitochondrial morphology
Changes in mitochondrial bioenergetics associated with LRRK2 G2019S
expression in fibroblasts and SHSY5Y cells have previously been shown to correlate
with mitochondrial morphological changes; mitochondria in LRRK2 G2019S patient
derived fibroblasts were described as elongated although no quantitative
assessment was carried out in the study [285]. In contrast mitochondrial
fragmentation was reported for primary neurons and SHSY5Y cells over-expressing
256
wild type and more so mutant LRRK2 [280, 281] where quantifications were limited
to neuronal processes with no calibration for the methodology. Visually,
mitochondria in cell bodies appeared un-fragmented when compared to stress
induced fragmentation in a number of other studies [135, 401]. Mitochondrial
fragmentation in the neuronal processes reported for LRRK2 mutants could
potentially relate to the mutant effects on neurite length and branching [259] more
so than the bioenergetic changes, an issue not addressed by either of the studies.
Although bioenergetic alterations were reported for both G2019S mutant
fibroblasts and SHSY5Y cells, mitochondrial uncoupling did not appear to perturb the
mitochondrial morphology. Intact mitochondrial networks described in patient
fibroblasts reported here and by Mortiboys et al suggest LRRK2 associated
mitochondrial fragmentation could result from LRRK2 over-expression or be limited
to neuronal models as fragmented structures were reported for both wild type and
mutant LRRK2 over-expressing SHSY5Y cells and primary cultures [280, 281].
Although difficult to assess without a direct comparison, the level of ectopic
expression of LRRK2 in the SHSY5Y model presented here may not be sufficient to
fragment the network.
In terms of a cell specific phenotype, comparable discrepancies have been
reported for studies involving PINK1. Mitochondrial fragmentation was reported in
fibroblasts derived from patients with familial PINK1 mutations, suggesting primary
skin cells are capable of this phenomenon [119]. In contrast genetic ablation of
PINK1 in Cos7 cells increased mitochondrial size through tubulation [402] while HeLa
and SHSY5Y PINK1 knockdown models presented with fragmented mitochondria
[119, 121]. Swollen mitochondria were described in human dopaminergic neurons
and primary mouse neuronal cultures from PINK1 knockout animals [403]. Electron
microscopy studies of recessive PD animal models revealed additional
257
inconsistencies where flies lacking PINK1 were reported to have swollen
mitochondria with disrupted cristae [404, 405] while genetic ablation in mice
produced only a mild mitochondrial phenotype with no obvious morphological
defects. Condensed and aggregated mitochondrial structures were also observed in
G2019S LRRK2 transgenic mice [233]. While Lin et al, reported mitochondria with
denser mitochondrial cristae in their G2019S LRRK2 transgenic mouse [235]. Animal
studies investigating mitochondrial morphology in PINK1 and LRRK2 PD associated
mutant did not reflect fragmented structures observed in the SHSY5Y cell models
[280, 281].
Although chemical uncouplers can induce mitochondrial fragmentation
[406], there is no evidence to suggest mild mitochondrial uncoupling correlates with
mitochondrial fragmentation. Uncoupled HtrA2 deficient mouse embryonic
fibroblasts and HeLa cells treated with HtrA2 siRNA showed elongated mitochondria
suggesting mitochondrial uncoupling in these cells correlated with mitochondrial
fusion [407]. UCP2-GFP over-expression in SHSY5Y cells did not appear to impact on
the mitochondrial morphology while mitochondrial uncoupling in G2019S expressing
fibroblasts and SHSY5Y cells did not correlate with mitochondrial fragmentation. In
addition to chemical uncouplers mitochondrial fragmentation can be induced by a
variety of cellular stressors such as nitric oxide (NO), FeSo4, N-methy-D-aspartic acid
(NMDA) agonists, extracellular β-amyloid aggregates [408] suggesting this process
may be a general feature of stressed cells. In support of this a genetic screen in
C.elegans showed that individual gene ablation of 80 % of all genes leads to
mitochondrial fragmentation [409]. Hence mitochondrial fragmentation is unlikely
to correlate with specific respiratory chain alterations or contribute to mitochondrial
uncoupling observed in G2019S expressing fibroblasts and SHSY5Y cells.
258
The bioenergetic changes and mitochondrial fragmentation associated with
mutant LRRK2 over-expression have been linked to increased cellular and
mitochondrial DLP1 levels [280, 281]. DLP1 mediated fragmentation has also been
described in the PINK1 and parkin PD models as well as for fragmented neuronal
cells following treatment with rotenone [410] suggesting defects in mitophagy and
Complex I inhibition can influence the fission/fusion machinery. The trigger for
mitochondrial DLP1 recruitment in mitochondrial toxin studies has been proposed to
be the defective electron transport chain. However, whether the mechanism is
membrane potential dependent or requires an enzyme stimulus has not been
addressed. A direct interaction between DLP1 and LRRK2 has been described [280,
281]. This association was enhanced for the G2019S mutant LRRK2 suggesting
mitochondrial LRRK2 could be the trigger for the DLP1 recruitment. A weak
association was detected for G2019S LRRK2 immunoprecipitated from SHSY5Y cells
absent for the wild type protein, potentially demonstrating enhanced mitochondrial
DLP1 recruitment by the mitochondrial G2019S LRRK2 pool. However, mitochondrial
LRRK2 and DLP1 abundance was comparable between control and G2019S
expressing SHSY5Y cells, in contrast to previous observations, suggesting increased
affinity for the G2019S LRRK2 DLP1 interaction which could take place in the cytosol,
a point not addressed in any of the studies. Kinase inhibition did not appear to affect
mitochondrial DLP1 recruitment further suggesting mitochondrial LRRK2 levels or
kinase activity did not influence DLP1 translocation in SHSY5Y cells. Our data
suggests that mitochondrial fragmentation and DLP1 levels are unlikely to contribute
to the mitochondrial uncoupling described in G2019S expressing fibroblasts and
SHSY5Y cells.
259
6.11 LRRK2, mitochondrial uncoupling and cellular disturbances implicated in PD
6.11.1 ATP, mitochondrial biogenesis and calcium
ATP depletion and bioenergetic insufficiency have been proposed to
contribute to cellular dysfunction and the death of dopaminergic neurons in PD.
Reduced ATP levels have been reported for PINK1 deficient mouse embryonic
fibroblasts [411], fibroblasts from patients with parkin mutations [346] and αsynuclein over-expression in HEK293 cells [412], correlating with respiratory chain
defects. Reduced ATP levels were also described in the HtrA2 deficient mouse
embryonic fibroblasts where mitochondrial uncoupling was described [413].
Reduced ATP levels have been described for G2019S expressing SHSY5Y cells [280]
and G2019S LRRK2 mutant fibroblasts [285]. Consistent with these observations, a
15 % reduction in basal ATP levels was reported for G2019S patient fibroblasts
relative to control cells described in this study suggesting reduced ATP levels are a
common feature of uncoupled cells and cells with respiratory chain defects,
potentially explaining some of the over-lapping pathology and clinical features
observed in genetically linked PD cases.
UCP levels have been shown to correlate with changes in cellular ATP and
mitochondrial biogenesis, a pathway associated with the recessively linked PD genes
suggesting dysregulated mitochondrial biogenesis may play a role in PD.
Downregulation of parkin correlated with reduced mitochondrial content [122]
while parkin over-expression was linked to enhanced mitochondrial biogenesis
[140]. Increased cellular ATP correlated with UCP1 expression in brown adipocytes
[414] correlating with increased mitochondrial biogenesis [415] while reduced
mitochondrial DNA content was linked to lower levels of UCP1 expression [416].
Increased UCP2 expression in muscle cells also enhanced mitochondrial biogenesis
and increased the cellular ATP content [373]. However, other studies have identified
260
mitochondrial uncoupling independent of mitochondrial biogenesis in UCP overexpressing systems; UCP2 over-expression resulted in mitochondrial uncoupling and
reduced basal ATP levels in pancreatic β-cells [417]. ATP levels were reduced in
skeletal muscle expressing high levels of UCP1 protein, relative to mice with lower
UCP1 expression [418]. As mitochondrial mass, respiratory chain subunit and TFAM
levels were comparable between control and G2019S LRRK2 expressing fibroblasts
we ruled out the potential contribution of mitochondrial biogenesis to mitochondrial
uncoupling in these cells.
As well as UCP regulation of the mitochondrial membrane potential and ATP
levels described in G2019S expressing fibroblasts and SHSY5Y cells, UCPs have a role
in the regulation of cellular calcium homeostasis. Calcium dysregulation is a noted
feature in PD; the loss of PINK1 protein has previously been shown to affect
mitochondrial calcium handling resulting in a decreased threshold for calcium
dependent cell death in primary neurons and neuroblastoma cells [172]. The
influence of UCPs on the mitochondrial membrane potential has previously been
related to their action as calcium uniporters, in line with the mechanism proposed to
regulate SNpc pacemaking [168]. In support of this, UCP2 and UCP3 over-expression
in human endothelial cells was associated with increased concentrations of free
mitochondrial calcium [419]. LRRK2 influence on UCP mRNA expression in fibroblasts
and SHSY5Y cells suggests that it is likely to play a role in the regulation of
mitochondrial calcium homeostasis which in turn may contribute to PD
pathogenesis.
6.11.2 LRRK2 regulation of autophagy and mitophagy
LRRK2 regulation of autophagy has been linked to the regulation of steady
state levels of misfolded proteins such as α-synuclein with the deregulation of this
pathway in G2019S expressing induced pluripotent dopaminergic stem cells
261
resulting in increased α-synuclein levels [420], potentially providing the link to the
pathological Lewy bodies observed in PD. However the mechanism of LRRK2
autophagic regulation is still under debate. Induced pluripotent stem cells generated
from PD patients with LRRK2 G2019S mutations were compromised at the level of
autophagic clearance [421] while autophagic flux was enhanced in G2019S patient
fibroblasts [422]. An increase in autophagic vacuoles was observed in differentiated
G2019S SHSY5Y cells [259] but whether these changes occurred as a result of
increased autophagic activation or reduced autophagic clearance was not
investigated. Tong et al, [240] reported a bi-phasic effect on autophagy in kidneys of
knockout animals where autophagic flux was enhanced at 7 months and perturbed
at 20 months suggesting age may contribute to the LRRK2 regulation of the
autophagy pathway and account for some of the discrepancies observed in the
experimental cell models. A number of studies emphasise unperturbed autophagic
flux associated with G2019S [271] and LRRK2 siRNA knockdown in T lymphocytes
and bone marrow derived macrophages [241]. The discrepancy may occur at the
tissue or cell source as differentiated neuronal cells, proliferating fibroblasts and
phagocytic immune cells have variable autophagic demands correlating with their
respective functional roles [423, 424]. A recent study has demonstrated that UCP2
over-expression in pancreatic adrenocarcinoma cells correlated with reduced
autophagy suggesting a potential link between mitochondrial uncoupling and the
autophagy cascade [425]. Whether these findings can be replicated in the SHSY5Y
system and extend to the UCP4 isoform warrants further investigation.
Although LRRK2 regulation of autophagy was not evaluated in G2019S
fibroblasts and SHSY5Y cells, mitochondrial removal by mitophagy was explored in
the context of LRRK2. Mitochondrial depolarization is thought to act as the trigger
for PINK1 mediated Parkin recruitment and stimulate mitophagy [426]. As the
262
mitochondrial membrane potential was reduced in G2019S expressing fibroblasts
and SHSY5Y cells, it was possible to speculate that excessive mitophagy may occur in
these cells. Although the process of mitochondrial turnover was not directly
evaluated, mitochondrial content and levels of respiratory chain subunits were
comparable between control and G2019S LRRK2 fibroblasts suggesting mitochondria
were not depleted. The levels of mitochondrial transcription factor, SOD2 and
catalase, genes that are part of the mitochondrial biosynthetic pathway were
comparable between control, wild type and G2019S expressing SHSY5Y cells
suggesting mitochondrial biosynthesis was not taking place. As mitochondrial
synthesis and number were comparable between control and G2019S cells,
excessive mitochondrial removal is unlikely to contribute to the mitochondrial
uncoupling observed in mutant fibroblasts and SHSY5Y cells.
6.11.3 LRRK2 regulation of synaptic function and tubulin dynamics
As well as mitochondrial morphology and autophagy, LRRK2 has been
implicated in the regulation of synaptic function [254, 255] and tubulin dynamics
[262, 427]. Mitochondria play a central role in meeting the demands of synapses for
ATP and regulating calcium homeostasis which in turn controls synaptic vesicle
endocytosis. UCP4 has previously been demonstrated to regulate calcium
homeostasis at synapses [428]. siRNA knockdown of LRRK2 in SHSY5Y correlated
with a 50 % reduction of UCP4 levels while LRRK2 silencing in presynaptic primary
cortical mouse neurons reduced synaptic transmission [254] and enhanced synaptic
vesicle recycling [255], both calcium dependent mechanisms [429, 430]. Overexpression of wild type LRRK2 impaired synaptic vesicle recycling in primary rat
hippocampal neurons, while the effect was more pronounced for the G2019S
mutant protein. The 5 fold increase in UCP4 mRNA in G2019S expressing SHSY5Y
263
cells is likely to result in reduced cytosolic calcium concentrations [419], which could
potentially account for the impaired vesicle recycling at the synapse [430].
Mitochondrial trafficking to and from the synapse requires the interaction
between glycogen synthase kinase 3β (GSK3β) and the Tau protein, forming
microtubules [431]. The speed of this process is dependent on GSK3β levels and the
subsequent phosphorylation of its Tau substrate [431]. Microtubule destabilization
correlates with reduced mitochondrial membrane potential and visa versa [432].
Hence UCP4 over-expression and the reduced mitochondrial membrane potential in
G2019S expressing SHSY5Y cells could impact on the mitochondrial association with
tubulin. In addition, the cellular and mitochondrial G2019S LRRK2 pools could
directly affect the phosphorylation status of tubulin which would in turn result in
reduced affinity of tubulin for mitochondria. In support of this LRRK2 has been
implicated in the regulation of microtubule assembly in mouse brain [262] with
interactions between LRRK2 and tubulin reported in mouse brain and HEK293 cells
[262]. Microtubule destabilization results in reduced mitochondrial trafficking to
synaptic terminals [433] reflecting how functional changes in the mitochondria could
affect mitochondrial delivery to the synapse and alter synaptic function in G2019S
expressing neuronal systems. In support of this, reduced mitochondrial numbers
have been noted in neuronal processes of G2019S expressing primary cortical
cultures and SHSY5Y cells [280, 281].
6.12 Mitochondrial uncoupling proteins and disease
Mitochondrial uncoupling has been associated with the loss of function of
the PD linked HtrA2 protein. Mouse embryonic fibroblasts generated from HtrA2
knockout animals had greater basal and oligomycin inhibited respiration rates
relative to their respective wild type counterparts [413] consistent with the
observations in G2019S LRRK2 expressing fibroblasts. The study reporting
264
mitochondrial uncoupling in HtrA2 knockout animals investigated the involvement
of UCP2 protein. Although the data was not shown, the authors claimed no changes
in UCP2 levels. The involvement of other UCP isoforms was not described hence the
involvement of these UCP family members in the regulation of HtrA2 knockout
mitochondrial phenotype cannot be ruled out. Alternative mechanisms for HtrA2
mediated mitochondrial uncoupling were not investigated. Although mitochondrial
uncoupling in disease is not common, increased respiration has also been described
for Luft’s disease. Although the etiology of the disorder is unknown the muscle
atrophy associated with the disease has been linked to hypermetabolism and
mitochondrial uncoupling in skeletal muscle [434, 435], potentially involving the
UCPs. HtrA2 linked PD and Luft’s disease emphasise the correlation between
mitochondrial uncoupling, neurodegeneration and skeletal pathology. In addition,
the genetic link between DJ1, recessively inherited PD and the UCP mediated
regulation of pacemaking activity specifically in the SNpc where cell death is
observed in PD patients further implicated the potential role of mitochondrial
uncoupling and UCPs in PD pathogenesis [168]. In support of this, pacemaker activity
involves the action of voltage dependent L type calcium channel with inhibitors
targeting this channel demonstrated to have a protective effect against PD in a
Danish cohort [436].
Mechanistically, mitochondrial uncoupling proteins have previously been
implicated in Parkinson’s disease as a protective mechanism against ROS [437-439].
Their protective effects have also been demonstrated on a background of
respiratory chain dysfunction models but the effects of chronic UCP upregulation
have not been investigated. In addition to protection from disease associated ROS,
UCP have been proposed to play a general role in ageing; over-expression of UCP5 in
Drosophila [440] correlated with increased lifespan, while UCP2 levels regulate
265
lifespan in mice [355]. However, the protective role of chronic UCP expression and
activation correlating with increased lifespan has been challenged. UCP activation by
fatty acids in fibroblasts was associated with premature senescence [441], as
assessed by β-galactosidase staining. Over-expression of UCP2 in COS7 cells had no
effect on lifespan but seemed to correlate with induction of apoptosis [167]. In
addition, although transient transfections of UCPs are well tolerated, prolonged
over-expression has been shown to result in death of HeLa, endothelial and
fibroblast cell lines [166]. The potential link of mitochondrial uncoupling and cell
death in PD has also been demonstrated by chronic exposure to low concentrations
of FCCP correlating with reduced lifespan in cell culture models as a result of
premature senescence [441]. Hepatic cells exposed to another uncoupler,
dinitrophenol retarded proliferation [442]. Although LRRK2 G2019S expression has
previously correlated with apoptotic induction [227, 286] senescence and cell death
were not observed in G2019S fibroblasts and SHSY5Y cells described here, however
compromised cell viability cannot be ruled out.
6.13 Endogenous LRRK2 regulation of UCPs
The G2019S mutation can either lead to a novel function or hyperactivate a
pathway regulated by wild type LRRK2. Two studies support the latter hypothesis;
siRNA knockdown of endogenous LRRK2 increased autophagic activity in HEK293
cells while expression of a BAC G2019S construct impaired autophagy [257]
suggesting this process is mediated by endogenous LRRK2 and dysregulated in the
presence of the mutant protein. The effects of wild type BAC LRRK2 expression were
not addressed. Ago2 mediated regulation of microRNA in Drosophila melanogaster
appeared to involve endogenous LRRK2 as siRNA knockdown of LRRK2 correlated
with reduced levels of dominant-negative dsRNA dependent protein kinase (DP1)
and eukaryotic translation initiation factor 2 alpha (EIF2α) proteins, the steady state
266
levels of which are regulated in a microRNA mediated fashion [273]. While wild type
LRRK2 over-expression in Drosophila melanogaster did not appear to grossly affect
DP1 or EIF2α levels, G2019S LRRK2 expression increased the levels of both of these
microRNA regulated proteins.
In support of these findings G2019S over-expression in SHSY5Y cells led to
mitochondrial uncoupling correlating with increased UCP4 levels. siRNA knockdown
of LRRK2 was associated with increased mitochondrial membrane potential and
reduced UCP4 expression suggesting endogenous LRRK2 in SHSY5Y cells may be
required for UCP4 transcriptional regulation. In line with the report by Gehrke et al,
wild type LRRK2 over-expression did not appear to impact on respiratory chain
function or UCP levels. If both mechanisms are dependent on the levels of the
associated LRRK2 substrate, increased levels of wild type protein would not affect
downstream signalling, reduced LRRK2 levels would reduce the LRRK2 substrate
interaction and downstream signalling while G2019S would compete with
endogenous LRRK2 for the substrate and the effects of increased LRRK2 kinase
activity could be implemented. This model is supported by LRRK2 kinase inhibitor
effects in control SHSY5Y cells where a mild effect on UCP2, 4 and 5 mRNA
expression was observed. UCP levels were unperturbed in wild type over-expressing
cells (kinase inhibition was not 100 % effective suggesting some active LRRK2
remained in the cells) while UCP4 levels in G2019S expressing cells were partially
restored. Therefore the G2019S mutation is likely to perturb LRRK2 regulation of
UCPs and impact on different aspects of cellular function, the cumulative effect of
these disturbances could contribute to PD pathogenesis.
267
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
Gasser T: Genetics of Parkinson's disease. Curr Opin Neurol 2005,
18(4):363-369.
Gibb WR: Functional neuropathology in Parkinson's disease. Eur
Neurol 1997, 38 Suppl 2:21-25.
Lanciego JL: Basal Ganglia Circuits: What's Now and Next?
Front Neuroanat 2012, 6:4.
Berardelli A, Rothwell JC, Thompson PD, Hallett M:
Pathophysiology of bradykinesia in Parkinson's disease. Brain
2001, 124(Pt 11):2131-2146.
Fahn S, Lang AE, Schapira AHV: Movement disorders 4.
Philadelphia, PA: Saunders/Elsevier; 2010.
Alexander GE, DeLong MR, Strick PL: Parallel organization of
functionally segregated circuits linking basal ganglia and cortex.
Annu Rev Neurosci 1986, 9:357-381.
Spillantini MG, Schmidt ML, Lee VM, Trojanowski JQ, Jakes R,
Goedert M: Alpha-synuclein in Lewy bodies. Nature 1997,
388(6645):839-840.
Blumberg MS, Freeman JH, Robinson SR: Oxford handbook of
developmental behavioral neuroscience. New York: Oxford
University Press; 2010.
Forno LS: Neuropathology of Parkinson's disease. J Neuropathol
Exp Neurol 1996, 55(3):259-272.
Kahle PJ: alpha-Synucleinopathy models and human
neuropathology: similarities and differences. Acta Neuropathol
2008, 115(1):87-95.
Elia AE, Lalli S, Albanese A: Differential diagnosis of dystonia.
Eur J Neurol 2010, 17 Suppl 1:1-8.
Wider C, Dickson DW, Wszolek ZK: Leucine-rich repeat kinase 2
gene-associated
disease:
redefining
genotype-phenotype
correlation. Neurodegener Dis 2010, 7(1-3):175-179.
Tan LC, Venketasubramanian N, Jamora RD, Heng D: Incidence
of Parkinson's disease in Singapore. Parkinsonism Relat Disord
2007, 13(1):40-43.
Polymeropoulos MH, Higgins JJ, Golbe LI, Johnson WG, Ide SE,
Di Iorio G, Sanges G, Stenroos ES, Pho LT, Schaffer AA et al:
Mapping of a gene for Parkinson's disease to chromosome 4q21q23. Science 1996, 274(5290):1197-1199.
Polymeropoulos MH, Lavedan C, Leroy E, Ide SE, Dehejia A,
Dutra A, Pike B, Root H, Rubenstein J, Boyer R et al: Mutation in
the alpha-synuclein gene identified in families with Parkinson's
disease. Science 1997, 276(5321):2045-2047.
Zarranz JJ, Alegre J, Gomez-Esteban JC, Lezcano E, Ros R,
Ampuero I, Vidal L, Hoenicka J, Rodriguez O, Atares B et al: The
new mutation, E46K, of alpha-synuclein causes Parkinson and
Lewy body dementia. Ann Neurol 2004, 55(2):164-173.
268
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
Chartier-Harlin MC, Kachergus J, Roumier C, Mouroux V,
Douay X, Lincoln S, Levecque C, Larvor L, Andrieux J, Hulihan
M et al: Alpha-synuclein locus duplication as a cause of familial
Parkinson's disease. Lancet 2004, 364(9440):1167-1169.
Singleton AB, Farrer M, Johnson J, Singleton A, Hague S,
Kachergus J, Hulihan M, Peuralinna T, Dutra A, Nussbaum R et
al: alpha-Synuclein locus triplication causes Parkinson's disease.
Science 2003, 302(5646):841.
Satake W, Nakabayashi Y, Mizuta I, Hirota Y, Ito C, Kubo M,
Kawaguchi T, Tsunoda T, Watanabe M, Takeda A et al: Genomewide association study identifies common variants at four loci as
genetic risk factors for Parkinson's disease. Nat Genet 2009,
41(12):1303-1307.
Simon-Sanchez J, Schulte C, Bras JM, Sharma M, Gibbs JR, Berg
D, Paisan-Ruiz C, Lichtner P, Scholz SW, Hernandez DG et al:
Genome-wide association study reveals genetic risk underlying
Parkinson's disease. Nat Genet 2009, 41(12):1308-1312.
Chiba-Falek O, Nussbaum RL: Effect of allelic variation at the
NACP-Rep1 repeat upstream of the alpha-synuclein gene (SNCA)
on transcription in a cell culture luciferase reporter system. Hum
Mol Genet 2001, 10(26):3101-3109.
Shin CW, Kim HJ, Park SS, Kim SY, Kim JY, Jeon BS: Two
Parkinson's disease patients with alpha-synuclein gene duplication
and rapid cognitive decline. Mov Disord 2010, 25(7):957-959.
Uversky VN: A protein-chameleon: conformational plasticity of
alpha-synuclein,
a
disordered
protein
involved
in
neurodegenerative disorders. J Biomol Struct Dyn 2003, 21(2):211234.
Iwai A, Masliah E, Yoshimoto M, Ge N, Flanagan L, de Silva HA,
Kittel A, Saitoh T: The precursor protein of non-A beta
component of Alzheimer's disease amyloid is a presynaptic protein
of the central nervous system. Neuron 1995, 14(2):467-475.
Kim HY, Heise H, Fernandez CO, Baldus M, Zweckstetter M:
Correlation of amyloid fibril beta-structure with the unfolded state
of alpha-synuclein. Chembiochem 2007, 8(14):1671-1674.
Davidson WS, Jonas A, Clayton DF, George JM: Stabilization of
alpha-synuclein secondary structure upon binding to synthetic
membranes. J Biol Chem 1998, 273(16):9443-9449.
Funayama M, Hasegawa K, Kowa H, Saito M, Tsuji S, Obata F: A
new locus for Parkinson's disease (PARK8) maps to chromosome
12p11.2-q13.1. Ann Neurol 2002, 51(3):296-301.
Zimprich A, Biskup S, Leitner P, Lichtner P, Farrer M, Lincoln S,
Kachergus J, Hulihan M, Uitti RJ, Calne DB et al: Mutations in
LRRK2
cause
autosomal-dominant
parkinsonism
with
pleomorphic pathology. Neuron 2004, 44(4):601-607.
Paisan-Ruiz C, Jain S, Evans EW, Gilks WP, Simon J, van der
Brug M, Lopez de Munain A, Aparicio S, Gil AM, Khan N et al:
Cloning of the gene containing mutations that cause PARK8linked Parkinson's disease. Neuron 2004, 44(4):595-600.
269
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
Hernandez D, Paisan Ruiz C, Crawley A, Malkani R, Werner J,
Gwinn-Hardy K, Dickson D, Wavrant Devrieze F, Hardy J,
Singleton A: The dardarin G 2019 S mutation is a common cause
of Parkinson's disease but not other neurodegenerative diseases.
Neurosci Lett 2005, 389(3):137-139.
Kachergus J, Mata IF, Hulihan M, Taylor JP, Lincoln S, Aasly J,
Gibson JM, Ross OA, Lynch T, Wiley J et al: Identification of a
novel LRRK2 mutation linked to autosomal dominant
parkinsonism: evidence of a common founder across European
populations. Am J Hum Genet 2005, 76(4):672-680.
Orr-Urtreger A, Shifrin C, Rozovski U, Rosner S, Bercovich D,
Gurevich T, Yagev-More H, Bar-Shira A, Giladi N: The LRRK2
G2019S mutation in Ashkenazi Jews with Parkinson disease: is
there a gender effect? Neurology 2007, 69(16):1595-1602.
Change N, Mercier G, Lucotte G: Genetic screening of the G2019S
mutation of the LRRK2 gene in Southwest European, North
African, and Sephardic Jewish subjects. Genet Test 2008,
12(3):333-339.
Gilks WP, Abou-Sleiman PM, Gandhi S, Jain S, Singleton A, Lees
AJ, Shaw K, Bhatia KP, Bonifati V, Quinn NP et al: A common
LRRK2 mutation in idiopathic Parkinson's disease. Lancet 2005,
365(9457):415-416.
Di Fonzo A, Rohe CF, Ferreira J, Chien HF, Vacca L, Stocchi F,
Guedes L, Fabrizio E, Manfredi M, Vanacore N et al: A frequent
LRRK2 gene mutation associated with autosomal dominant
Parkinson's disease. Lancet 2005, 365(9457):412-415.
Hulihan MM, Ishihara-Paul L, Kachergus J, Warren L, Amouri
R, Elango R, Prinjha RK, Upmanyu R, Kefi M, Zouari M et al:
LRRK2 Gly2019Ser penetrance in Arab-Berber patients from
Tunisia: a case-control genetic study. Lancet Neurol 2008,
7(7):591-594.
Zabetian CP, Morino H, Ujike H, Yamamoto M, Oda M,
Maruyama H, Izumi Y, Kaji R, Griffith A, Leis BC et al:
Identification and haplotype analysis of LRRK2 G2019S in
Japanese patients with Parkinson disease. Neurology 2006,
67(4):697-699.
Greggio E, Lewis PA, van der Brug MP, Ahmad R, Kaganovich A,
Ding J, Beilina A, Baker AK, Cookson MR: Mutations in
LRRK2/dardarin associated with Parkinson disease are more toxic
than equivalent mutations in the homologous kinase LRRK1. J
Neurochem 2007, 102(1):93-102.
Gaig C, Marti MJ, Ezquerra M, Rey MJ, Cardozo A, Tolosa E:
G2019S LRRK2 mutation causing Parkinson's disease without
Lewy bodies. J Neurol Neurosurg Psychiatry 2007, 78(6):626-628.
Ross OA, Toft M, Whittle AJ, Johnson JL, Papapetropoulos S,
Mash DC, Litvan I, Gordon MF, Wszolek ZK, Farrer MJ et al:
Lrrk2 and Lewy body disease. Ann Neurol 2006, 59(2):388-393.
Khan NL, Jain S, Lynch JM, Pavese N, Abou-Sleiman P, Holton
JL, Healy DG, Gilks WP, Sweeney MG, Ganguly M et al:
Mutations in the gene LRRK2 encoding dardarin (PARK8) cause
270
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
familial Parkinson's disease: clinical, pathological, olfactory and
functional imaging and genetic data. Brain 2005, 128(Pt 12):27862796.
Kitada T, Asakawa S, Hattori N, Matsumine H, Yamamura Y,
Minoshima S, Yokochi M, Mizuno Y, Shimizu N: Mutations in the
parkin gene cause autosomal recessive juvenile parkinsonism.
Nature 1998, 392(6676):605-608.
Lucking CB, Durr A, Bonifati V, Vaughan J, De Michele G,
Gasser T, Harhangi BS, Meco G, Denefle P, Wood NW et al:
Association between early-onset Parkinson's disease and mutations
in the parkin gene. N Engl J Med 2000, 342(21):1560-1567.
Pramstaller PP, Schlossmacher MG, Jacques TS, Scaravilli F,
Eskelson C, Pepivani I, Hedrich K, Adel S, Gonzales-McNeal M,
Hilker R et al: Lewy body Parkinson's disease in a large pedigree
with 77 Parkin mutation carriers. Ann Neurol 2005, 58(3):411-422.
Houlden H, Singleton AB: The genetics and neuropathology of
Parkinson's disease. Acta Neuropathol 2012, 124(3):325-338.
Valente EM, Abou-Sleiman PM, Caputo V, Muqit MM, Harvey K,
Gispert S, Ali Z, Del Turco D, Bentivoglio AR, Healy DG et al:
Hereditary early-onset Parkinson's disease caused by mutations in
PINK1. Science 2004, 304(5674):1158-1160.
Rogaeva E, Johnson J, Lang AE, Gulick C, Gwinn-Hardy K,
Kawarai T, Sato C, Morgan A, Werner J, Nussbaum R et al:
Analysis of the PINK1 gene in a large cohort of cases with
Parkinson disease. Arch Neurol 2004, 61(12):1898-1904.
Valente EM, Brancati F, Caputo V, Graham EA, Davis MB,
Ferraris A, Breteler MM, Gasser T, Bonifati V, Bentivoglio AR et
al: PARK6 is a common cause of familial parkinsonism. Neurol Sci
2002, 23 Suppl 2:S117-118.
Valente EM, Salvi S, Ialongo T, Marongiu R, Elia AE, Caputo V,
Romito L, Albanese A, Dallapiccola B, Bentivoglio AR: PINK1
mutations are associated with sporadic early-onset parkinsonism.
Ann Neurol 2004, 56(3):336-341.
Samaranch L, Lorenzo-Betancor O, Arbelo JM, Ferrer I, Lorenzo
E, Irigoyen J, Pastor MA, Marrero C, Isla C, Herrera-Henriquez J
et al: PINK1-linked parkinsonism is associated with Lewy body
pathology. Brain 2010, 133(Pt 4):1128-1142.
Greene AW, Grenier K, Aguileta MA, Muise S, Farazifard R,
Haque ME, McBride HM, Park DS, Fon EA: Mitochondrial
processing peptidase regulates PINK1 processing, import and
Parkin recruitment. EMBO Rep 2012, 13(4):378-385.
Bonifati V, Rizzu P, van Baren MJ, Schaap O, Breedveld GJ,
Krieger E, Dekker MC, Squitieri F, Ibanez P, Joosse M et al:
Mutations in the DJ-1 gene associated with autosomal recessive
early-onset parkinsonism. Science 2003, 299(5604):256-259.
Canet-Aviles RM, Wilson MA, Miller DW, Ahmad R, McLendon
C, Bandyopadhyay S, Baptista MJ, Ringe D, Petsko GA, Cookson
MR: The Parkinson's disease protein DJ-1 is neuroprotective due
to cysteine-sulfinic acid-driven mitochondrial localization. Proc
Natl Acad Sci U S A 2004, 101(24):9103-9108.
271
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
Sidransky E, Nalls MA, Aasly JO, Aharon-Peretz J, Annesi G,
Barbosa ER, Bar-Shira A, Berg D, Bras J, Brice A et al:
Multicenter analysis of glucocerebrosidase mutations in
Parkinson's disease. N Engl J Med 2009, 361(17):1651-1661.
Aharon-Peretz J, Rosenbaum H, Gershoni-Baruch R: Mutations
in the glucocerebrosidase gene and Parkinson's disease in
Ashkenazi Jews. N Engl J Med 2004, 351(19):1972-1977.
Lwin A, Orvisky E, Goker-Alpan O, LaMarca ME, Sidransky E:
Glucocerebrosidase mutations in subjects with parkinsonism. Mol
Genet Metab 2004, 81(1):70-73.
Nichols WC, Pankratz N, Marek DK, Pauciulo MW, Elsaesser VE,
Halter CA, Rudolph A, Wojcieszek J, Pfeiffer RF, Foroud T:
Mutations in GBA are associated with familial Parkinson disease
susceptibility and age at onset. Neurology 2009, 72(4):310-316.
Mitsui J, Mizuta I, Toyoda A, Ashida R, Takahashi Y, Goto J,
Fukuda Y, Date H, Iwata A, Yamamoto M et al: Mutations for
Gaucher disease confer high susceptibility to Parkinson disease.
Arch Neurol 2009, 66(5):571-576.
Neumann J, Bras J, Deas E, O'Sullivan SS, Parkkinen L,
Lachmann RH, Li A, Holton J, Guerreiro R, Paudel R et al:
Glucocerebrosidase mutations in clinical and pathologically
proven Parkinson's disease. Brain 2009, 132(Pt 7):1783-1794.
Chen M, Wang J: Gaucher disease: review of the literature. Arch
Pathol Lab Med 2008, 132(5):851-853.
Sawkar AR, Adamski-Werner SL, Cheng WC, Wong CH, Beutler
E, Zimmer KP, Kelly JW: Gaucher disease-associated
glucocerebrosidases
show
mutation-dependent
chemical
chaperoning profiles. Chem Biol 2005, 12(11):1235-1244.
Masliah E, Rockenstein E, Veinbergs I, Mallory M, Hashimoto M,
Takeda A, Sagara Y, Sisk A, Mucke L: Dopaminergic loss and
inclusion body formation in alpha-synuclein mice: implications for
neurodegenerative disorders. Science 2000, 287(5456):1265-1269.
Neumann M, Kahle PJ, Giasson BI, Ozmen L, Borroni E, Spooren
W, Muller V, Odoy S, Fujiwara H, Hasegawa M et al: Misfolded
proteinase K-resistant hyperphosphorylated alpha-synuclein in
aged transgenic mice with locomotor deterioration and in human
alpha-synucleinopathies. J Clin Invest 2002, 110(10):1429-1439.
Giasson BI, Duda JE, Quinn SM, Zhang B, Trojanowski JQ, Lee
VM: Neuronal alpha-synucleinopathy with severe movement
disorder in mice expressing A53T human alpha-synuclein. Neuron
2002, 34(4):521-533.
Richfield EK, Thiruchelvam MJ, Cory-Slechta DA, Wuertzer C,
Gainetdinov RR, Caron MG, Di Monte DA, Federoff HJ:
Behavioral and neurochemical effects of wild-type and mutated
human alpha-synuclein in transgenic mice. Exp Neurol 2002,
175(1):35-48.
Nuber S, Petrasch-Parwez E, Winner B, Winkler J, von Horsten S,
Schmidt T, Boy J, Kuhn M, Nguyen HP, Teismann P et al:
Neurodegeneration and motor dysfunction in a conditional model
of Parkinson's disease. J Neurosci 2008, 28(10):2471-2484.
272
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
Lee MK, Stirling W, Xu Y, Xu X, Qui D, Mandir AS, Dawson TM,
Copeland NG, Jenkins NA, Price DL: Human alpha-synucleinharboring familial Parkinson's disease-linked Ala-53 --> Thr
mutation causes neurodegenerative disease with alpha-synuclein
aggregation in transgenic mice. Proc Natl Acad Sci U S A 2002,
99(13):8968-8973.
Feany MB, Bender WW: A Drosophila model of Parkinson's
disease. Nature 2000, 404(6776):394-398.
Takahashi M, Kanuka H, Fujiwara H, Koyama A, Hasegawa M,
Miura M, Iwatsubo T: Phosphorylation of alpha-synuclein
characteristic of synucleinopathy lesions is recapitulated in alphasynuclein transgenic Drosophila. Neurosci Lett 2003, 336(3):155158.
Kuwahara T, Koyama A, Gengyo-Ando K, Masuda M, Kowa H,
Tsunoda M, Mitani S, Iwatsubo T: Familial Parkinson mutant
alpha-synuclein causes dopamine neuron dysfunction in transgenic
Caenorhabditis elegans. J Biol Chem 2006, 281(1):334-340.
Lakso M, Vartiainen S, Moilanen AM, Sirvio J, Thomas JH, Nass
R, Blakely RD, Wong G: Dopaminergic neuronal loss and motor
deficits in Caenorhabditis elegans overexpressing human alphasynuclein. J Neurochem 2003, 86(1):165-172.
Goldberg MS, Fleming SM, Palacino JJ, Cepeda C, Lam HA,
Bhatnagar A, Meloni EG, Wu N, Ackerson LC, Klapstein GJ et al:
Parkin-deficient mice exhibit nigrostriatal deficits but not loss of
dopaminergic neurons. J Biol Chem 2003, 278(44):43628-43635.
Itier JM, Ibanez P, Mena MA, Abbas N, Cohen-Salmon C, Bohme
GA, Laville M, Pratt J, Corti O, Pradier L et al: Parkin gene
inactivation alters behaviour and dopamine neurotransmission in
the mouse. Hum Mol Genet 2003, 12(18):2277-2291.
Von Coelln R, Thomas B, Savitt JM, Lim KL, Sasaki M, Hess EJ,
Dawson VL, Dawson TM: Loss of locus coeruleus neurons and
reduced startle in parkin null mice. Proc Natl Acad Sci U S A 2004,
101(29):10744-10749.
Greene JC, Whitworth AJ, Kuo I, Andrews LA, Feany MB,
Pallanck LJ: Mitochondrial pathology and apoptotic muscle
degeneration in Drosophila parkin mutants. Proc Natl Acad Sci U
S A 2003, 100(7):4078-4083.
Whitworth AJ, Theodore DA, Greene JC, Benes H, Wes PD,
Pallanck LJ: Increased glutathione S-transferase activity rescues
dopaminergic neuron loss in a Drosophila model of Parkinson's
disease. Proc Natl Acad Sci U S A 2005, 102(22):8024-8029.
Hao LY, Giasson BI, Bonini NM: DJ-1 is critical for mitochondrial
function and rescues PINK1 loss of function. Proc Natl Acad Sci U
S A 2010, 107(21):9747-9752.
Gispert S, Ricciardi F, Kurz A, Azizov M, Hoepken HH, Becker D,
Voos W, Leuner K, Muller WE, Kudin AP et al: Parkinson
phenotype in aged PINK1-deficient mice is accompanied by
progressive
mitochondrial
dysfunction
in
absence
of
neurodegeneration. PLoS One 2009, 4(6):e5777.
273
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
91.
92.
Rugbjerg K, Ritz B, Korbo L, Martinussen N, Olsen JH: Risk of
Parkinson's disease after hospital contact for head injury:
population based case-control study. BMJ 2008, 337:a2494.
Hernan MA, Takkouche B, Caamano-Isorna F, Gestal-Otero JJ: A
meta-analysis of coffee drinking, cigarette smoking, and the risk of
Parkinson's disease. Ann Neurol 2002, 52(3):276-284.
Quik M, O'Neill M, Perez XA: Nicotine neuroprotection against
nigrostriatal damage: importance of the animal model. Trends
Pharmacol Sci 2007, 28(5):229-235.
Frigerio R, Elbaz A, Sanft KR, Peterson BJ, Bower JH, Ahlskog
JE, Grossardt BR, de Andrade M, Maraganore DM, Rocca WA:
Education and occupations preceding Parkinson disease: a
population-based case-control study. Neurology 2005, 65(10):15751583.
Saaksjarvi K, Knekt P, Rissanen H, Laaksonen MA, Reunanen A,
Mannisto S: Prospective study of coffee consumption and risk of
Parkinson's disease. Eur J Clin Nutr 2008, 62(7):908-915.
McGeer PL, Itagaki S, Boyes BE, McGeer EG: Reactive microglia
are positive for HLA-DR in the substantia nigra of Parkinson's
and Alzheimer's disease brains. Neurology 1988, 38(8):1285-1291.
Langston JW, Ballard P, Tetrud JW, Irwin I: Chronic
Parkinsonism in humans due to a product of meperidine-analog
synthesis. Science 1983, 219(4587):979-980.
Tanner CM, Kamel F, Ross GW, Hoppin JA, Goldman SM, Korell
M, Marras C, Bhudhikanok GS, Kasten M, Chade AR et al:
Rotenone, paraquat, and Parkinson's disease. Environ Health
Perspect 2011, 119(6):866-872.
Porras G, Li Q, Bezard E: Modeling Parkinson's Disease in
Primates: The MPTP Model. Cold Spring Harb Perspect Med 2012,
2(3):a009308.
Betarbet R, Sherer TB, MacKenzie G, Garcia-Osuna M, Panov
AV, Greenamyre JT: Chronic systemic pesticide exposure
reproduces features of Parkinson's disease. Nat Neurosci 2000,
3(12):1301-1306.
Brooks AI, Chadwick CA, Gelbard HA, Cory-Slechta DA,
Federoff HJ: Paraquat elicited neurobehavioral syndrome caused
by dopaminergic neuron loss. Brain Res 1999, 823(1-2):1-10.
McCormack AL, Thiruchelvam M, Manning-Bog AB, Thiffault C,
Langston JW, Cory-Slechta DA, Di Monte DA: Environmental
risk factors and Parkinson's disease: selective degeneration of
nigral dopaminergic neurons caused by the herbicide paraquat.
Neurobiol Dis 2002, 10(2):119-127.
Cuperus R, Leen R, Tytgat GA, Caron HN, van Kuilenburg AB:
Fenretinide induces mitochondrial ROS and inhibits the
mitochondrial respiratory chain in neuroblastoma. Cell Mol Life
Sci 2010, 67(5):807-816.
Turrens JF: Mitochondrial formation of reactive oxygen species. J
Physiol 2003, 552(Pt 2):335-344.
274
93.
94.
95.
96.
97.
98.
99.
100.
101.
102.
103.
104.
105.
106.
Chiba K, Trevor AJ, Castagnoli N, Jr.: Active uptake of MPP+, a
metabolite of MPTP, by brain synaptosomes. Biochem Biophys Res
Commun 1985, 128(3):1228-1232.
Nicklas WJ, Vyas I, Heikkila RE: Inhibition of NADH-linked
oxidation in brain mitochondria by 1-methyl-4-phenyl-pyridine, a
metabolite of the neurotoxin, 1-methyl-4-phenyl-1,2,5,6tetrahydropyridine. Life Sci 1985, 36(26):2503-2508.
Richardson JR, Quan Y, Sherer TB, Greenamyre JT, Miller GW:
Paraquat neurotoxicity is distinct from that of MPTP and
rotenone. Toxicol Sci 2005, 88(1):193-201.
Bonneh-Barkay D, Langston WJ, Di Monte DA: Toxicity of redox
cycling pesticides in primary mesencephalic cultures. Antioxid
Redox Signal 2005, 7(5-6):649-653.
Mann VM, Cooper JM, Krige D, Daniel SE, Schapira AH,
Marsden CD: Brain, skeletal muscle and platelet homogenate
mitochondrial function in Parkinson's disease. Brain 1992, 115 ( Pt
2):333-342.
Schapira AH, Cooper JM, Dexter D, Jenner P, Clark JB, Marsden
CD: Mitochondrial complex I deficiency in Parkinson's disease.
Lancet 1989, 1(8649):1269.
Schapira AH, Cooper JM, Dexter D, Clark JB, Jenner P, Marsden
CD: Mitochondrial complex I deficiency in Parkinson's disease. J
Neurochem 1990, 54(3):823-827.
Janetzky B, Hauck S, Youdim MB, Riederer P, Jellinger K,
Pantucek F, Zochling R, Boissl KW, Reichmann H: Unaltered
aconitase activity, but decreased complex I activity in substantia
nigra pars compacta of patients with Parkinson's disease. Neurosci
Lett 1994, 169(1-2):126-128.
Keeney PM, Xie J, Capaldi RA, Bennett JP, Jr.: Parkinson's
disease brain mitochondrial complex I has oxidatively damaged
subunits and is functionally impaired and misassembled. J
Neurosci 2006, 26(19):5256-5264.
Bindoff LA, Birch-Machin MA, Cartlidge NE, Parker WD, Jr.,
Turnbull DM: Respiratory chain abnormalities in skeletal muscle
from patients with Parkinson's disease. J Neurol Sci 1991,
104(2):203-208.
Shinde S, Pasupathy K: Respiratory-chain enzyme activities in
isolated mitochondria of lymphocytes from patients with
Parkinson's disease: preliminary study. Neurol India 2006,
54(4):390-393.
Yoshino H, Nakagawa-Hattori Y, Kondo T, Mizuno Y:
Mitochondrial complex I and II activities of lymphocytes and
platelets in Parkinson's disease. J Neural Transm Park Dis Dement
Sect 1992, 4(1):27-34.
Haas RH, Nasirian F, Nakano K, Ward D, Pay M, Hill R, Shults
CW: Low platelet mitochondrial complex I and complex II/III
activity in early untreated Parkinson's disease. Ann Neurol 1995,
37(6):714-722.
Benecke R, Strumper P, Weiss H: Electron transfer complexes I
and IV of platelets are abnormal in Parkinson's disease but
275
107.
108.
109.
110.
111.
112.
113.
114.
115.
116.
117.
normal in Parkinson-plus syndromes. Brain 1993, 116 ( Pt 6):14511463.
Dexter DT, Carter CJ, Wells FR, Javoy-Agid F, Agid Y, Lees A,
Jenner P, Marsden CD: Basal lipid peroxidation in substantia
nigra is increased in Parkinson's disease. J Neurochem 1989,
52(2):381-389.
Alam ZI, Daniel SE, Lees AJ, Marsden DC, Jenner P, Halliwell B:
A generalised increase in protein carbonyls in the brain in
Parkinson's but not incidental Lewy body disease. J Neurochem
1997, 69(3):1326-1329.
Floor E, Wetzel MG: Increased protein oxidation in human
substantia nigra pars compacta in comparison with basal ganglia
and prefrontal cortex measured with an improved
dinitrophenylhydrazine assay. J Neurochem 1998, 70(1):268-275.
Nakabeppu Y, Tsuchimoto D, Yamaguchi H, Sakumi K: Oxidative
damage in nucleic acids and Parkinson's disease. J Neurosci Res
2007, 85(5):919-934.
Selley ML: (E)-4-hydroxy-2-nonenal may be involved in the
pathogenesis of Parkinson's disease. Free Radic Biol Med 1998,
25(2):169-174.
Kikuchi A, Takeda A, Onodera H, Kimpara T, Hisanaga K, Sato
N, Nunomura A, Castellani RJ, Perry G, Smith MA et al: Systemic
increase of oxidative nucleic acid damage in Parkinson's disease
and multiple system atrophy. Neurobiol Dis 2002, 9(2):244-248.
Abe T, Isobe C, Murata T, Sato C, Tohgi H: Alteration of 8hydroxyguanosine concentrations in the cerebrospinal fluid and
serum from patients with Parkinson's disease. Neurosci Lett 2003,
336(2):105-108.
Buhmann C, Arlt S, Kontush A, Moller-Bertram T, Sperber S,
Oechsner M, Stuerenburg HJ, Beisiegel U: Plasma and CSF
markers of oxidative stress are increased in Parkinson's disease
and influenced by antiparkinsonian medication. Neurobiol Dis
2004, 15(1):160-170.
Krebiehl G, Ruckerbauer S, Burbulla LF, Kieper N, Maurer B,
Waak J, Wolburg H, Gizatullina Z, Gellerich FN, Woitalla D et al:
Reduced basal autophagy and impaired mitochondrial dynamics
due to loss of Parkinson's disease-associated protein DJ-1. PLoS
One 2010, 5(2):e9367.
Thomas KJ, McCoy MK, Blackinton J, Beilina A, van der Brug M,
Sandebring A, Miller D, Maric D, Cedazo-Minguez A, Cookson
MR: DJ-1 acts in parallel to the PINK1/parkin pathway to control
mitochondrial function and autophagy. Hum Mol Genet 2011,
20(1):40-50.
Heo JY, Park JH, Kim SJ, Seo KS, Han JS, Lee SH, Kim JM, Park
JI, Park SK, Lim K et al: DJ-1 null dopaminergic neuronal cells
exhibit defects in mitochondrial function and structure:
involvement of mitochondrial complex I assembly. PLoS One 2012,
7(3):e32629.
276
118.
119.
120.
121.
122.
123.
124.
125.
126.
127.
128.
129.
130.
131.
Palacino JJ, Sagi D, Goldberg MS, Krauss S, Motz C, Wacker M,
Klose J, Shen J: Mitochondrial dysfunction and oxidative damage
in parkin-deficient mice. J Biol Chem 2004, 279(18):18614-18622.
Exner N, Treske B, Paquet D, Holmstrom K, Schiesling C, Gispert
S, Carballo-Carbajal I, Berg D, Hoepken HH, Gasser T et al: Lossof-function of human PINK1 results in mitochondrial pathology
and can be rescued by parkin. J Neurosci 2007, 27(45):1241312418.
Hoepken HH, Gispert S, Morales B, Wingerter O, Del Turco D,
Mulsch A, Nussbaum RL, Muller K, Drose S, Brandt U et al:
Mitochondrial dysfunction, peroxidation damage and changes in
glutathione metabolism in PARK6. Neurobiol Dis 2007, 25(2):401411.
Dagda RK, Cherra SJ, 3rd, Kulich SM, Tandon A, Park D, Chu
CT: Loss of PINK1 function promotes mitophagy through effects
on oxidative stress and mitochondrial fission. J Biol Chem 2009,
284(20):13843-13855.
Gegg ME, Cooper JM, Schapira AH, Taanman JW: Silencing of
PINK1 expression affects mitochondrial DNA and oxidative
phosphorylation in dopaminergic cells. PLoS One 2009, 4(3):e4756.
Devi L, Raghavendran V, Prabhu BM, Avadhani NG,
Anandatheerthavarada HK: Mitochondrial
import and
accumulation of alpha-synuclein impair complex I in human
dopaminergic neuronal cultures and Parkinson disease brain. J
Biol Chem 2008, 283(14):9089-9100.
Chinta SJ, Mallajosyula JK, Rane A, Andersen JK: Mitochondrial
alpha-synuclein accumulation impairs complex I function in
dopaminergic neurons and results in increased mitophagy in vivo.
Neurosci Lett 2010, 486(3):235-239.
Suen DF, Narendra DP, Tanaka A, Manfredi G, Youle RJ: Parkin
overexpression selects against a deleterious mtDNA mutation in
heteroplasmic cybrid cells. Proc Natl Acad Sci U S A 2010,
107(26):11835-11840.
Youle RJ, van der Bliek AM: Mitochondrial fission, fusion, and
stress. Science 2012, 337(6098):1062-1065.
Baker MJ, Tatsuta T, Langer T: Quality control of mitochondrial
proteostasis. Cold Spring Harb Perspect Biol 2011, 3(7).
Narendra D, Tanaka A, Suen DF, Youle RJ: Parkin is recruited
selectively to impaired mitochondria and promotes their
autophagy. J Cell Biol 2008, 183(5):795-803.
Gegg ME, Cooper JM, Chau KY, Rojo M, Schapira AH, Taanman
JW: Mitofusin 1 and mitofusin 2 are ubiquitinated in a
PINK1/parkin-dependent manner upon induction of mitophagy.
Hum Mol Genet 2010, 19(24):4861-4870.
Narendra D, Tanaka A, Suen DF, Youle RJ: Parkin-induced
mitophagy in the pathogenesis of Parkinson disease. Autophagy
2009, 5(5):706-708.
Jin SM, Lazarou M, Wang C, Kane LA, Narendra DP, Youle RJ:
Mitochondrial membrane potential regulates PINK1 import and
277
132.
133.
134.
135.
136.
137.
138.
139.
140.
141.
142.
143.
144.
145.
proteolytic destabilization by PARL. J Cell Biol 2010, 191(5):933942.
Narendra DP, Jin SM, Tanaka A, Suen DF, Gautier CA, Shen J,
Cookson MR, Youle RJ: PINK1 is selectively stabilized on
impaired mitochondria to activate Parkin. PLoS Biol 2010,
8(1):e1000298.
Liu W, Acin-Perez R, Geghman KD, Manfredi G, Lu B, Li C:
Pink1 regulates the oxidative phosphorylation machinery via
mitochondrial fission. Proc Natl Acad Sci U S A 2011,
108(31):12920-12924.
Ron I, Rapaport D, Horowitz M: Interaction between parkin and
mutant glucocerebrosidase variants: a possible link between
Parkinson disease and Gaucher disease. Hum Mol Genet 2010,
19(19):3771-3781.
Kamp F, Exner N, Lutz AK, Wender N, Hegermann J, Brunner B,
Nuscher B, Bartels T, Giese A, Beyer K et al: Inhibition of
mitochondrial fusion by alpha-synuclein is rescued by PINK1,
Parkin and DJ-1. EMBO J 2010, 29(20):3571-3589.
Nakamura K, Nemani VM, Azarbal F, Skibinski G, Levy JM,
Egami K, Munishkina L, Zhang J, Gardner B, Wakabayashi J et
al: Direct membrane association drives mitochondrial fission by
the Parkinson disease-associated protein alpha-synuclein. J Biol
Chem 2011, 286(23):20710-20726.
Kobayashi M, Yamamoto M: Molecular mechanisms activating
the Nrf2-Keap1 pathway of antioxidant gene regulation. Antioxid
Redox Signal 2005, 7(3-4):385-394.
Sykiotis GP, Bohmann D: Keap1/Nrf2 signaling regulates
oxidative stress tolerance and lifespan in Drosophila. Dev Cell
2008, 14(1):76-85.
Piantadosi CA, Suliman HB: Mitochondrial transcription factor A
induction by redox activation of nuclear respiratory factor 1. J
Biol Chem 2006, 281(1):324-333.
Kuroda Y, Mitsui T, Kunishige M, Shono M, Akaike M, Azuma H,
Matsumoto T: Parkin enhances mitochondrial biogenesis in
proliferating cells. Hum Mol Genet 2006, 15(6):883-895.
Valle I, Alvarez-Barrientos A, Arza E, Lamas S, Monsalve M:
PGC-1alpha regulates the mitochondrial antioxidant defense
system in vascular endothelial cells. Cardiovasc Res 2005,
66(3):562-573.
Liang H, Ward WF: PGC-1alpha: a key regulator of energy
metabolism. Adv Physiol Educ 2006, 30(4):145-151.
Li W, Kong AN: Molecular mechanisms of Nrf2-mediated
antioxidant response. Mol Carcinog 2009, 48(2):91-104.
Czubryt MP, McAnally J, Fishman GI, Olson EN: Regulation of
peroxisome proliferator-activated receptor gamma coactivator 1
alpha (PGC-1 alpha ) and mitochondrial function by MEF2 and
HDAC5. Proc Natl Acad Sci U S A 2003, 100(4):1711-1716.
Zhong N, Xu J: Synergistic activation of the human MnSOD
promoter by DJ-1 and PGC-1alpha: regulation by SUMOylation
and oxidation. Hum Mol Genet 2008, 17(21):3357-3367.
278
146.
147.
148.
149.
150.
151.
152.
153.
154.
155.
156.
157.
158.
Junn E, Jang WH, Zhao X, Jeong BS, Mouradian MM:
Mitochondrial localization of DJ-1 leads to enhanced
neuroprotection. J Neurosci Res 2009, 87(1):123-129.
Xu J, Zhong N, Wang H, Elias JE, Kim CY, Woldman I, Pifl C,
Gygi SP, Geula C, Yankner BA: The Parkinson's diseaseassociated DJ-1 protein is a transcriptional co-activator that
protects against neuronal apoptosis. Hum Mol Genet 2005,
14(9):1231-1241.
Castillo-Quan JI: Parkin' control: regulation of PGC-1alpha
through PARIS in Parkinson's disease. Dis Model Mech 2011,
4(4):427-429.
Shin JH, Ko HS, Kang H, Lee Y, Lee YI, Pletinkova O, Troconso
JC, Dawson VL, Dawson TM: PARIS (ZNF746) repression of
PGC-1alpha contributes to neurodegeneration in Parkinson's
disease. Cell 2011, 144(5):689-702.
Sadeghian M, Marinova-Mutafchieva L, Broom L, Davis JB,
Virley D, Medhurst AD, Dexter DT: Full and partial peroxisome
proliferation-activated receptor-gamma agonists, but not delta
agonist, rescue of dopaminergic neurons in the 6-OHDA
parkinsonian model is associated with inhibition of microglial
activation and MMP expression. J Neuroimmunol 2012, 246(12):69-77.
Friling S, Bergsland M, Kjellander S: Activation of Retinoid X
Receptor increases dopamine cell survival in models for
Parkinson's disease. BMC Neurosci 2009, 10:146.
Clark J, Reddy S, Zheng K, Betensky RA, Simon DK: Association
of PGC-1alpha polymorphisms with age of onset and risk of
Parkinson's disease. BMC Med Genet 2011, 12:69.
Ledesma A, de Lacoba MG, Rial E: The mitochondrial uncoupling
proteins. Genome Biol 2002, 3(12):REVIEWS3015.
Ricquier D, Bouillaud F: The uncoupling protein homologues:
UCP1, UCP2, UCP3, StUCP and AtUCP. Biochem J 2000, 345 Pt
2:161-179.
Porter RK: Mitochondrial proton leak: a role for uncoupling
proteins 2 and 3? Biochim Biophys Acta 2001, 1504(1):120-127.
Medvedev AV, Snedden SK, Raimbault S, Ricquier D, Collins S:
Transcriptional regulation of the mouse uncoupling protein-2
gene. Double E-box motif is required for peroxisome proliferatoractivated receptor-gamma-dependent activation. J Biol Chem
2001, 276(14):10817-10823.
Young ME, Patil S, Ying J, Depre C, Ahuja HS, Shipley GL,
Stepkowski SM, Davies PJ, Taegtmeyer H: Uncoupling protein 3
transcription is regulated by peroxisome proliferator-activated
receptor (alpha) in the adult rodent heart. FASEB J 2001,
15(3):833-845.
Pecqueur C, Alves-Guerra MC, Gelly C, Levi-Meyrueis C,
Couplan E, Collins S, Ricquier D, Bouillaud F, Miroux B:
Uncoupling protein 2, in vivo distribution, induction upon
oxidative stress, and evidence for translational regulation. J Biol
Chem 2001, 276(12):8705-8712.
279
159.
160.
161.
162.
163.
164.
165.
166.
167.
168.
169.
170.
172.
173.
Carroll AM, Porter RK, Morrice NA: Identification of serine
phosphorylation in mitochondrial uncoupling protein 1. Biochim
Biophys Acta 2008, 1777(7-8):1060-1065.
Azzu V, Affourtit C, Breen EP, Parker N, Brand MD: Dynamic
regulation of uncoupling protein 2 content in INS-1E insulinoma
cells. Biochim Biophys Acta 2008, 1777(10):1378-1383.
Derdak Z, Fulop P, Sabo E, Tavares R, Berthiaume EP, Resnick
MB, Paragh G, Wands JR, Baffy G: Enhanced colon tumor
induction in uncoupling protein-2 deficient mice is associated with
NF-kappaB activation and oxidative stress. Carcinogenesis 2006,
27(5):956-961.
Shilo S, Aharoni-Simon M, Tirosh O: Selenium attenuates
expression of MnSOD and uncoupling protein 2 in J774.2
macrophages: molecular mechanism for its cell-death and
antiinflammatory activity. Antioxid Redox Signal 2005, 7(1-2):276286.
Jaburek M, Miyamoto S, Di Mascio P, Garlid KD, Jezek P:
Hydroperoxy fatty acid cycling mediated by mitochondrial
uncoupling protein UCP2. J Biol Chem 2004, 279(51):53097-53102.
Ho PW, Chan DY, Kwok KH, Chu AC, Ho JW, Kung MH,
Ramsden DB, Ho SL: Methyl-4-phenylpyridinium ion modulates
expression of mitochondrial uncoupling proteins 2, 4, and 5 in
catecholaminergic (SK-N-SH) cells. J Neurosci Res 2005,
81(2):261-268.
Echtay KS, Pakay JL, Esteves TC, Brand MD: Hydroxynonenal
and uncoupling proteins: a model for protection against oxidative
damage. Biofactors 2005, 24(1-4):119-130.
Dikov D, Aulbach A, Muster B, Drose S, Jendrach M, BereiterHahn J: Do UCP2 and mild uncoupling improve longevity? Exp
Gerontol 2010, 45(7-8):586-595.
McDonald RB, Walker KM, Warman DB, Griffey SM, Warden
CH, Ramsey JJ, Horwitz BA: Characterization of survival and
phenotype throughout the life span in UCP2/UCP3 genetically
altered mice. Exp Gerontol 2008, 43(12):1061-1068.
Guzman JN, Sanchez-Padilla J, Wokosin D, Kondapalli J, Ilijic E,
Schumacker PT, Surmeier DJ: Oxidant stress evoked by
pacemaking in dopaminergic neurons is attenuated by DJ-1.
Nature 2010, 468(7324):696-700.
Albin RL, Young AB, Penney JB: The functional anatomy of basal
ganglia disorders. Trends Neurosci 1989, 12(10):366-375.
Lipscombe D: L-type calcium channels: highs and new lows. Circ
Res 2002, 90(9):933-935.
Gandhi S, Wood-Kaczmar A, Yao Z, Plun-Favreau H, Deas E,
Klupsch K, Downward J, Latchman DS, Tabrizi SJ, Wood NW et
al: PINK1-associated Parkinson's disease is caused by neuronal
vulnerability to calcium-induced cell death. Mol Cell 2009,
33(5):627-638.
Frei B, Richter C: N-methyl-4-phenylpyridine (MMP+) together
with 6-hydroxydopamine or dopamine stimulates Ca2+ release
from mitochondria. FEBS Lett 1986, 198(1):99-102.
280
174.
175.
176.
177.
178.
179.
180.
181.
182.
183.
184.
185.
186.
Sousa SC, Maciel EN, Vercesi AE, Castilho RF: Ca2+-induced
oxidative stress in brain mitochondria treated with the respiratory
chain inhibitor rotenone. FEBS Lett 2003, 543(1-3):179-183.
Wang XJ, Xu JX: Possible involvement of Ca2+ signaling in
rotenone-induced apoptosis in human neuroblastoma SH-SY5Y
cells. Neurosci Lett 2005, 376(2):127-132.
Vila M, Vukosavic S, Jackson-Lewis V, Neystat M, Jakowec M,
Przedborski S: Alpha-synuclein up-regulation in substantia nigra
dopaminergic neurons following administration of the
parkinsonian toxin MPTP. J Neurochem 2000, 74(2):721-729.
Meredith GE, Totterdell S, Petroske E, Santa Cruz K, Callison
RC, Jr., Lau YS: Lysosomal malfunction accompanies alphasynuclein aggregation in a progressive mouse model of Parkinson's
disease. Brain Res 2002, 956(1):156-165.
Yu WH, Matsuoka Y, Sziraki I, Hashim A, Lafrancois J, Sershen
H, Duff KE: Increased dopaminergic neuron sensitivity to 1methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) in transgenic
mice expressing mutant A53T alpha-synuclein. Neurochem Res
2008, 33(5):902-911.
Nieto M, Gil-Bea FJ, Dalfo E, Cuadrado M, Cabodevilla F,
Sanchez B, Catena S, Sesma T, Ribe E, Ferrer I et al: Increased
sensitivity to MPTP in human alpha-synuclein A30P transgenic
mice. Neurobiol Aging 2006, 27(6):848-856.
McLean JR, Hallett PJ, Cooper O, Stanley M, Isacson O:
Transcript expression levels of full-length alpha-synuclein and its
three alternatively spliced variants in Parkinson's disease brain
regions and in a transgenic mouse model of alpha-synuclein
overexpression. Mol Cell Neurosci 2012, 49(2):230-239.
St Martin JL, Klucken J, Outeiro TF, Nguyen P, Keller-McGandy
C, Cantuti-Castelvetri I, Grammatopoulos TN, Standaert DG,
Hyman BT, McLean PJ: Dopaminergic neuron loss and upregulation of chaperone protein mRNA induced by targeted overexpression of alpha-synuclein in mouse substantia nigra. J
Neurochem 2007, 100(6):1449-1457.
Cuervo AM, Stefanis L, Fredenburg R, Lansbury PT, Sulzer D:
Impaired degradation of mutant alpha-synuclein by chaperonemediated autophagy. Science 2004, 305(5688):1292-1295.
Sridhar S, Botbol Y, Macian F, Cuervo AM: Autophagy and
disease: always two sides to a problem. J Pathol 2012, 226(2):255273.
Yamamoto A, Cremona ML, Rothman JE: Autophagy-mediated
clearance of huntingtin aggregates triggered by the insulinsignaling pathway. J Cell Biol 2006, 172(5):719-731.
Berger Z, Ravikumar B, Menzies FM, Oroz LG, Underwood BR,
Pangalos MN, Schmitt I, Wullner U, Evert BO, O'Kane CJ et al:
Rapamycin alleviates toxicity of different aggregate-prone
proteins. Hum Mol Genet 2006, 15(3):433-442.
Sarkar S, Floto RA, Berger Z, Imarisio S, Cordenier A, Pasco M,
Cook LJ, Rubinsztein DC: Lithium induces autophagy by
281
187.
188.
189.
190.
191.
192.
193.
194.
195.
196.
197.
inhibiting inositol monophosphatase. J Cell Biol 2005, 170(7):11011111.
Ravikumar B, Vacher C, Berger Z, Davies JE, Luo S, Oroz LG,
Scaravilli F, Easton DF, Duden R, O'Kane CJ et al: Inhibition of
mTOR induces autophagy and reduces toxicity of polyglutamine
expansions in fly and mouse models of Huntington disease. Nat
Genet 2004, 36(6):585-595.
Xilouri M, Vogiatzi T, Vekrellis K, Park D, Stefanis L: Abberant
alpha-synuclein confers toxicity to neurons in part through
inhibition of chaperone-mediated autophagy. PLoS One 2009,
4(5):e5515.
Qiao L, Hamamichi S, Caldwell KA, Caldwell GA, Yacoubian TA,
Wilson S, Xie ZL, Speake LD, Parks R, Crabtree D et al:
Lysosomal enzyme cathepsin D protects against alpha-synuclein
aggregation and toxicity. Mol Brain 2008, 1:17.
Wong K, Sidransky E, Verma A, Mixon T, Sandberg GD,
Wakefield LK, Morrison A, Lwin A, Colegial C, Allman JM et al:
Neuropathology provides clues to the pathophysiology of Gaucher
disease. Mol Genet Metab 2004, 82(3):192-207.
Saito Y, Ruberu NN, Sawabe M, Arai T, Kazama H, Hosoi T,
Yamanouchi H, Murayama S: Lewy body-related alphasynucleinopathy in aging. J Neuropathol Exp Neurol 2004,
63(7):742-749.
Cullen V, Lindfors M, Ng J, Paetau A, Swinton E, Kolodziej P,
Boston H, Saftig P, Woulfe J, Feany MB et al: Cathepsin D
expression level affects alpha-synuclein processing, aggregation,
and toxicity in vivo. Mol Brain 2009, 2:5.
Goker-Alpan O, Giasson BI, Eblan MJ, Nguyen J, Hurtig HI, Lee
VM, Trojanowski JQ, Sidransky E: Glucocerebrosidase mutations
are an important risk factor for Lewy body disorders. Neurology
2006, 67(5):908-910.
Mazzulli JR, Xu YH, Sun Y, Knight AL, McLean PJ, Caldwell
GA, Sidransky E, Grabowski GA, Krainc D: Gaucher disease
glucocerebrosidase and alpha-synuclein form a bidirectional
pathogenic loop in synucleinopathies. Cell 2011, 146(1):37-52.
Sardi SP, Clarke J, Kinnecom C, Tamsett TJ, Li L, Stanek LM,
Passini MA, Grabowski GA, Schlossmacher MG, Sidman RL et al:
CNS expression of glucocerebrosidase corrects alpha-synuclein
pathology and memory in a mouse model of Gaucher-related
synucleinopathy. Proc Natl Acad Sci U S A 2011, 108(29):1210112106.
Dehay B, Ramirez A, Martinez-Vicente M, Perier C, Canron MH,
Doudnikoff E, Vital A, Vila M, Klein C, Bezard E: Loss of P-type
ATPase ATP13A2/PARK9 function induces general lysosomal
deficiency and leads to Parkinson disease neurodegeneration. Proc
Natl Acad Sci U S A 2012, 109(24):9611-9616.
Usenovic M, Tresse E, Mazzulli JR, Taylor JP, Krainc D:
Deficiency of ATP13A2 leads to lysosomal dysfunction, alphasynuclein accumulation, and neurotoxicity. J Neurosci 2012,
32(12):4240-4246.
282
198.
199.
200.
201.
202.
203.
204.
205.
206.
208.
209.
210.
211.
Jolly RD, Brown S, Das AM, Walkley SU: Mitochondrial
dysfunction in the neuronal ceroid-lipofuscinoses (Batten disease).
Neurochem Int 2002, 40(6):565-571.
Luiro K, Kopra O, Blom T, Gentile M, Mitchison HM, Hovatta I,
Tornquist K, Jalanko A: Batten disease (JNCL) is linked to
disturbances in mitochondrial, cytoskeletal, and synaptic
compartments. J Neurosci Res 2006, 84(5):1124-1138.
Lee BD, Dawson VL, Dawson TM: Leucine-rich repeat kinase 2
(LRRK2) as a potential therapeutic target in Parkinson's disease.
Trends Pharmacol Sci 2012, 33(7):365-373.
West AB, Moore DJ, Biskup S, Bugayenko A, Smith WW, Ross
CA, Dawson VL, Dawson TM: Parkinson's disease-associated
mutations in leucine-rich repeat kinase 2 augment kinase activity.
Proc Natl Acad Sci U S A 2005, 102(46):16842-16847.
Jaleel M, Nichols RJ, Deak M, Campbell DG, Gillardon F, Knebel
A, Alessi DR: LRRK2 phosphorylates moesin at threonine-558:
characterization of how Parkinson's disease mutants affect kinase
activity. Biochem J 2007, 405(2):307-317.
West AB, Moore DJ, Choi C, Andrabi SA, Li X, Dikeman D,
Biskup S, Zhang Z, Lim KL, Dawson VL et al: Parkinson's
disease-associated mutations in LRRK2 link enhanced GTPbinding and kinase activities to neuronal toxicity. Hum Mol Genet
2007, 16(2):223-232.
Ito G, Okai T, Fujino G, Takeda K, Ichijo H, Katada T, Iwatsubo
T: GTP binding is essential to the protein kinase activity of
LRRK2, a causative gene product for familial Parkinson's disease.
Biochemistry 2007, 46(5):1380-1388.
Liu M, Dobson B, Glicksman MA, Yue Z, Stein RL: Kinetic
mechanistic studies of wild-type leucine-rich repeat kinase 2:
characterization of the kinase and GTPase activities. Biochemistry
2010, 49(9):2008-2017.
Taymans JM, Vancraenenbroeck R, Ollikainen P, Beilina A,
Lobbestael E, De Maeyer M, Baekelandt V, Cookson MR: LRRK2
kinase activity is dependent on LRRK2 GTP binding capacity but
independent of LRRK2 GTP binding. PLoS One 2011, 6(8):e23207.
Jorgensen ND, Peng Y, Ho CC, Rideout HJ, Petrey D, Liu P,
Dauer WT: The WD40 domain is required for LRRK2
neurotoxicity. PLoS One 2009, 4(12):e8463.
Deng J, Lewis PA, Greggio E, Sluch E, Beilina A, Cookson MR:
Structure of the ROC domain from the Parkinson's diseaseassociated leucine-rich repeat kinase 2 reveals a dimeric GTPase.
Proc Natl Acad Sci U S A 2008, 105(5):1499-1504.
Klein CL, Rovelli G, Springer W, Schall C, Gasser T, Kahle PJ:
Homo- and heterodimerization of ROCO kinases: LRRK2 kinase
inhibition by the LRRK2 ROCO fragment. J Neurochem 2009,
111(3):703-715.
Greggio E, Taymans JM, Zhen EY, Ryder J, Vancraenenbroeck
R, Beilina A, Sun P, Deng J, Jaffe H, Baekelandt V et al: The
Parkinson's disease kinase LRRK2 autophosphorylates its GTPase
283
212.
213.
214.
215.
216.
217.
218.
219.
220.
221.
222.
223.
domain at multiple sites. Biochem Biophys Res Commun 2009,
389(3):449-454.
Deng X, Dzamko N, Prescott A, Davies P, Liu Q, Yang Q, Lee JD,
Patricelli MP, Nomanbhoy TK, Alessi DR et al: Characterization
of a selective inhibitor of the Parkinson's disease kinase LRRK2.
Nat Chem Biol 2011, 7(4):203-205.
Ramsden N, Perrin J, Ren Z, Lee BD, Zinn N, Dawson VL, Tam
D, Bova M, Lang M, Drewes G et al: Chemoproteomics-based
design of potent LRRK2-selective lead compounds that attenuate
Parkinson's disease-related toxicity in human neurons. ACS Chem
Biol 2011, 6(10):1021-1028.
Zhang J, Deng X, Choi HG, Alessi DR, Gray NS: Characterization
of TAE684 as a potent LRRK2 kinase inhibitor. Bioorg Med Chem
Lett 2012, 22(5):1864-1869.
Nichols RJ, Dzamko N, Morrice NA, Campbell DG, Deak M,
Ordureau A, Macartney T, Tong Y, Shen J, Prescott AR et al: 143-3 binding to LRRK2 is disrupted by multiple Parkinson's
disease-associated
mutations
and
regulates
cytoplasmic
localization. Biochem J 2010, 430(3):393-404.
Dzamko N, Deak M, Hentati F, Reith AD, Prescott AR, Alessi DR,
Nichols RJ: Inhibition of LRRK2 kinase activity leads to
dephosphorylation of Ser(910)/Ser(935), disruption of 14-3-3
binding and altered cytoplasmic localization. Biochem J 2010,
430(3):405-413.
Di Fonzo A, Tassorelli C, De Mari M, Chien HF, Ferreira J, Rohe
CF, Riboldazzi G, Antonini A, Albani G, Mauro A et al:
Comprehensive analysis of the LRRK2 gene in sixty families with
Parkinson's disease. Eur J Hum Genet 2006, 14(3):322-331.
Gloeckner CJ, Kinkl N, Schumacher A, Braun RJ, O'Neill E,
Meitinger T, Kolch W, Prokisch H, Ueffing M: The Parkinson
disease causing LRRK2 mutation I2020T is associated with
increased kinase activity. Hum Mol Genet 2006, 15(2):223-232.
Lewis PA, Greggio E, Beilina A, Jain S, Baker A, Cookson MR:
The R1441C mutation of LRRK2 disrupts GTP hydrolysis.
Biochem Biophys Res Commun 2007, 357(3):668-671.
Daniels V, Vancraenenbroeck R, Law BM, Greggio E, Lobbestael
E, Gao F, De Maeyer M, Cookson MR, Harvey K, Baekelandt V et
al: Insight into the mode of action of the LRRK2 Y1699C
pathogenic mutant. J Neurochem 2011, 116(2):304-315.
Greggio E, Zambrano I, Kaganovich A, Beilina A, Taymans JM,
Daniels V, Lewis P, Jain S, Ding J, Syed A et al: The Parkinson
disease-associated leucine-rich repeat kinase 2 (LRRK2) is a dimer
that undergoes intramolecular autophosphorylation. J Biol Chem
2008, 283(24):16906-16914.
Smith WW, Pei Z, Jiang H, Dawson VL, Dawson TM, Ross CA:
Kinase activity of mutant LRRK2 mediates neuronal toxicity. Nat
Neurosci 2006, 9(10):1231-1233.
Smith WW, Pei Z, Jiang H, Moore DJ, Liang Y, West AB, Dawson
VL, Dawson TM, Ross CA: Leucine-rich repeat kinase 2 (LRRK2)
284
224.
225.
226.
227.
228.
229.
230.
231.
232.
233.
234.
235.
236.
interacts with parkin, and mutant LRRK2 induces neuronal
degeneration. Proc Natl Acad Sci U S A 2005, 102(51):18676-18681.
Greggio E, Jain S, Kingsbury A, Bandopadhyay R, Lewis P,
Kaganovich A, van der Brug MP, Beilina A, Blackinton J, Thomas
KJ et al: Kinase activity is required for the toxic effects of mutant
LRRK2/dardarin. Neurobiol Dis 2006, 23(2):329-341.
Xiong H, Qiu H, Zhuang L, Jiang R, Chen Y: Effects of 5-AzaCdR on the proliferation of human breast cancer cell line MCF-7
and on the expression of Apaf-1 gene. J Huazhong Univ Sci
Technolog Med Sci 2009, 29(4):498-502.
Lichtenberg M, Mansilla A, Zecchini VR, Fleming A, Rubinsztein
DC: The Parkinson's disease protein LRRK2 impairs proteasome
substrate clearance without affecting proteasome catalytic activity.
Cell Death Dis 2011, 2:e196.
Iaccarino C, Crosio C, Vitale C, Sanna G, Carri MT, Barone P:
Apoptotic mechanisms in mutant LRRK2-mediated cell death.
Hum Mol Genet 2007, 16(11):1319-1326.
Sen S, Webber PJ, West AB: Dependence of leucine-rich repeat
kinase 2 (LRRK2) kinase activity on dimerization. J Biol Chem
2009, 284(52):36346-36356.
Samann J, Hegermann J, von Gromoff E, Eimer S, Baumeister R,
Schmidt E: Caenorhabditits elegans LRK-1 and PINK-1 act
antagonistically in stress response and neurite outgrowth. J Biol
Chem 2009, 284(24):16482-16491.
Wolozin B, Saha S, Guillily M, Ferree A, Riley M: Investigating
convergent actions of genes linked to familial Parkinson's disease.
Neurodegener Dis 2008, 5(3-4):182-185.
Imai Y, Gehrke S, Wang HQ, Takahashi R, Hasegawa K, Oota E,
Lu B: Phosphorylation of 4E-BP by LRRK2 affects the
maintenance of dopaminergic neurons in Drosophila. EMBO J
2008, 27(18):2432-2443.
Liu Z, Wang X, Yu Y, Li X, Wang T, Jiang H, Ren Q, Jiao Y,
Sawa A, Moran T et al: A Drosophila model for LRRK2-linked
parkinsonism. Proc Natl Acad Sci U S A 2008, 105(7):2693-2698.
Ramonet D, Daher JP, Lin BM, Stafa K, Kim J, Banerjee R,
Westerlund M, Pletnikova O, Glauser L, Yang L et al:
Dopaminergic neuronal loss, reduced neurite complexity and
autophagic abnormalities in transgenic mice expressing G2019S
mutant LRRK2. PLoS One 2011, 6(4):e18568.
Li X, Patel JC, Wang J, Avshalumov MV, Nicholson C, Buxbaum
JD, Elder GA, Rice ME, Yue Z: Enhanced striatal dopamine
transmission and motor performance with LRRK2 overexpression
in mice is eliminated by familial Parkinson's disease mutation
G2019S. J Neurosci 2010, 30(5):1788-1797.
Lin X, Parisiadou L, Gu XL, Wang L, Shim H, Sun L, Xie C, Long
CX, Yang WJ, Ding J et al: Leucine-rich repeat kinase 2 regulates
the progression of neuropathology induced by Parkinson's-diseaserelated mutant alpha-synuclein. Neuron 2009, 64(6):807-827.
Daher JP, Pletnikova O, Biskup S, Musso A, Gellhaar S, Galter D,
Troncoso JC, Lee MK, Dawson TM, Dawson VL et al:
285
237.
238.
239.
240.
241.
242.
243.
244.
245.
246.
247.
248.
Neurodegenerative phenotypes in an A53T alpha-synuclein
transgenic mouse model are independent of LRRK2. Hum Mol
Genet 2012, 21(11):2420-2431.
Biskup S, Moore DJ, Rea A, Lorenz-Deperieux B, Coombes CE,
Dawson VL, Dawson TM, West AB: Dynamic and redundant
regulation of LRRK2 and LRRK1 expression. BMC Neurosci 2007,
8:102.
Herzig MC, Kolly C, Persohn E, Theil D, Schweizer T, Hafner T,
Stemmelen C, Troxler TJ, Schmid P, Danner S et al: LRRK2
protein levels are determined by kinase function and are crucial
for kidney and lung homeostasis in mice. Hum Mol Genet 2011,
20(21):4209-4223.
Hatano T, Kubo S, Imai S, Maeda M, Ishikawa K, Mizuno Y,
Hattori N: Leucine-rich repeat kinase 2 associates with lipid rafts.
Hum Mol Genet 2007, 16(6):678-690.
Tong Y, Giaime E, Yamaguchi H, Ichimura T, Liu Y, Si H, Cai H,
Bonventre JV, Shen J: Loss of leucine-rich repeat kinase 2 causes
age-dependent bi-phasic alterations of the autophagy pathway.
Mol Neurodegener 2012, 7:2.
Hakimi M, Selvanantham T, Swinton E, Padmore RF, Tong Y,
Kabbach G, Venderova K, Girardin SE, Bulman DE, Scherzer CR
et al: Parkinson's disease-linked LRRK2 is expressed in circulating
and tissue immune cells and upregulated following recognition of
microbial structures. J Neural Transm 2011, 118(5):795-808.
Kubo M, Kamiya Y, Nagashima R, Maekawa T, Eshima K, Azuma
S, Ohta E, Obata F: LRRK2 is expressed in B-2 but not in B-1 B
cells, and downregulated by cellular activation. J Neuroimmunol
2010, 229(1-2):123-128.
Gardet A, Benita Y, Li C, Sands BE, Ballester I, Stevens C,
Korzenik JR, Rioux JD, Daly MJ, Xavier RJ et al: LRRK2 is
involved in the IFN-gamma response and host response to
pathogens. J Immunol 2010, 185(9):5577-5585.
Melrose HL, Kent CB, Taylor JP, Dachsel JC, Hinkle KM, Lincoln
SJ, Mok SS, Culvenor JG, Masters CL, Tyndall GM et al: A
comparative analysis of leucine-rich repeat kinase 2 (Lrrk2)
expression in mouse brain and Lewy body disease. Neuroscience
2007, 147(4):1047-1058.
Melrose H, Lincoln S, Tyndall G, Dickson D, Farrer M:
Anatomical localization of leucine-rich repeat kinase 2 in mouse
brain. Neuroscience 2006, 139(3):791-794.
Simon-Sanchez J, Herranz-Perez V, Olucha-Bordonau F, PerezTur J: LRRK2 is expressed in areas affected by Parkinson's
disease in the adult mouse brain. Eur J Neurosci 2006, 23(3):659666.
Higashi S, Biskup S, West AB, Trinkaus D, Dawson VL, Faull RL,
Waldvogel HJ, Arai H, Dawson TM, Moore DJ et al: Localization
of Parkinson's disease-associated LRRK2 in normal and
pathological human brain. Brain Res 2007, 1155:208-219.
Higashi S, Moore DJ, Colebrooke RE, Biskup S, Dawson VL, Arai
H, Dawson TM, Emson PC: Expression and localization of
286
249.
250.
251.
252.
253.
254.
255.
256.
257.
258.
259.
260.
261.
Parkinson's disease-associated leucine-rich repeat kinase 2 in the
mouse brain. J Neurochem 2007, 100(2):368-381.
Lee H, Melrose HL, Yue M, Pare JF, Farrer MJ, Smith Y: Lrrk2
localization in the primate basal ganglia and thalamus: a light and
electron microscopic analysis in monkeys. Exp Neurol 2010,
224(2):438-447.
Mandemakers W, Snellinx A, O'Neill MJ, de Strooper B: LRRK2
expression is enriched in the striosomal compartment of mouse
striatum. Neurobiol Dis 2012, 48(3):582-593.
Biskup S, Moore DJ, Celsi F, Higashi S, West AB, Andrabi SA,
Kurkinen K, Yu SW, Savitt JM, Waldvogel HJ et al: Localization
of LRRK2 to membranous and vesicular structures in mammalian
brain. Ann Neurol 2006, 60(5):557-569.
Berger Z, Smith KA, Lavoie MJ: Membrane localization of
LRRK2 is associated with increased formation of the highly active
LRRK2 dimer and changes in its phosphorylation. Biochemistry
2010, 49(26):5511-5523.
Lee S, Liu HP, Lin WY, Guo H, Lu B: LRRK2 kinase regulates
synaptic morphology through distinct substrates at the presynaptic
and postsynaptic compartments of the Drosophila neuromuscular
junction. J Neurosci 2010, 30(50):16959-16969.
Piccoli G, Condliffe SB, Bauer M, Giesert F, Boldt K, De Astis S,
Meixner A, Sarioglu H, Vogt-Weisenhorn DM, Wurst W et al:
LRRK2 controls synaptic vesicle storage and mobilization within
the recycling pool. J Neurosci 2011, 31(6):2225-2237.
Shin N, Jeong H, Kwon J, Heo HY, Kwon JJ, Yun HJ, Kim CH,
Han BS, Tong Y, Shen J et al: LRRK2 regulates synaptic vesicle
endocytosis. Exp Cell Res 2008, 314(10):2055-2065.
Lundblad M, Decressac M, Mattsson B, Bjorklund A: Impaired
neurotransmission caused by overexpression of alpha-synuclein in
nigral dopamine neurons. Proc Natl Acad Sci U S A 2012,
109(9):3213-3219.
Alegre-Abarrategui J, Christian H, Lufino MM, Mutihac R,
Venda LL, Ansorge O, Wade-Martins R: LRRK2 regulates
autophagic activity and localizes to specific membrane
microdomains in a novel human genomic reporter cellular model.
Hum Mol Genet 2009, 18(21):4022-4034.
Heo HY, Kim KS, Seol W: Coordinate Regulation of Neurite
Outgrowth by LRRK2 and Its Interactor, Rab5. Exp Neurobiol
2010, 19(2):97-105.
Plowey ED, Cherra SJ, 3rd, Liu YJ, Chu CT: Role of autophagy in
G2019S-LRRK2-associated neurite shortening in differentiated
SH-SY5Y cells. J Neurochem 2008, 105(3):1048-1056.
MacLeod D, Dowman J, Hammond R, Leete T, Inoue K,
Abeliovich A: The familial Parkinsonism gene LRRK2 regulates
neurite process morphology. Neuron 2006, 52(4):587-593.
Gillardon F: Interaction of elongation factor 1-alpha with leucinerich repeat kinase 2 impairs kinase activity and microtubule
bundling in vitro. Neuroscience 2009, 163(2):533-539.
287
262.
263.
264.
265.
266.
267.
268.
269.
270.
271.
272.
273.
274.
275.
Gillardon F: Leucine-rich repeat kinase 2 phosphorylates brain
tubulin-beta isoforms and modulates microtubule stability--a point
of convergence in parkinsonian neurodegeneration? J Neurochem
2009, 110(5):1514-1522.
Kawakami F, Yabata T, Ohta E, Maekawa T, Shimada N, Suzuki
M, Maruyama H, Ichikawa T, Obata F: LRRK2 phosphorylates
tubulin-associated tau but not the free molecule: LRRK2-mediated
regulation of the tau-tubulin association and neurite outgrowth.
PLoS One 2012, 7(1):e30834.
Sancho RM, Law BM, Harvey K: Mutations in the LRRK2 RocCOR tandem domain link Parkinson's disease to Wnt signalling
pathways. Hum Mol Genet 2009, 18(20):3955-3968.
Duka T, Duka V, Joyce JN, Sidhu A: Alpha-Synuclein contributes
to GSK-3beta-catalyzed Tau phosphorylation in Parkinson's
disease models. FASEB J 2009, 23(9):2820-2830.
Habig K, Walter M, Poths S, Riess O, Bonin M: RNA interference
of LRRK2-microarray expression analysis of a Parkinson's disease
key player. Neurogenetics 2008, 9(2):83-94.
Schulz C, Paus M, Frey K, Schmid R, Kohl Z, Mennerich D,
Winkler J, Gillardon F: Leucine-rich repeat kinase 2 modulates
retinoic acid-induced neuronal differentiation of murine
embryonic stem cells. PLoS One 2011, 6(6):e20820.
Chiba-Falek O, Lopez GJ, Nussbaum RL: Levels of alphasynuclein mRNA in sporadic Parkinson disease patients. Mov
Disord 2006, 21(10):1703-1708.
Liu GH, Qu J, Suzuki K, Nivet E, Li M, Montserrat N, Yi F, Xu X,
Ruiz S, Zhang W et al: Progressive degeneration of human neural
stem cells caused by pathogenic LRRK2. Nature 2012,
491(7425):603-607.
Carballo-Carbajal I, Weber-Endress S, Rovelli G, Chan D,
Wolozin B, Klein CL, Patenge N, Gasser T, Kahle PJ: Leucinerich repeat kinase 2 induces alpha-synuclein expression via the
extracellular signal-regulated kinase pathway. Cell Signal 2010,
22(5):821-827.
Liu Z, Lee J, Krummey S, Lu W, Cai H, Lenardo MJ: The kinase
LRRK2 is a regulator of the transcription factor NFAT that
modulates the severity of inflammatory bowel disease. Nat
Immunol 2011, 12(11):1063-1070.
Berwick DC, Harvey K: LRRK2 functions as a Wnt signaling
scaffold, bridging cytosolic proteins and membrane-localized
LRP6. Hum Mol Genet 2012, 21(22):4966-4979.
Gehrke S, Imai Y, Sokol N, Lu B: Pathogenic LRRK2 negatively
regulates microRNA-mediated translational repression. Nature
2010, 466(7306):637-641.
Pons B, Armengol G, Livingstone M, Lopez L, Coch L, Sonenberg
N, Ramon y Cajal S: Association between LRRK2 and 4E-BP1
protein levels in normal and malignant cells. Oncol Rep 2012,
27(1):225-231.
Tain LS, Mortiboys H, Tao RN, Ziviani E, Bandmann O,
Whitworth AJ: Rapamycin activation of 4E-BP prevents
288
276.
277.
278.
279.
280.
281.
282.
283.
284.
285.
286.
287.
288.
parkinsonian dopaminergic neuron loss. Nat Neurosci 2009,
12(9):1129-1135.
Chen CY, Weng YH, Chien KY, Lin KJ, Yeh TH, Cheng YP, Lu
CS, Wang HL: (G2019S) LRRK2 activates MKK4-JNK pathway
and causes degeneration of SN dopaminergic neurons in a
transgenic mouse model of PD. Cell Death Differ 2012,
19(10):1623-1633.
White LR, Toft M, Kvam SN, Farrer MJ, Aasly JO: MAPKpathway activity, Lrrk2 G2019S, and Parkinson's disease. J
Neurosci Res 2007, 85(6):1288-1294.
Hunot S, Brugg B, Ricard D, Michel PP, Muriel MP, Ruberg M,
Faucheux BA, Agid Y, Hirsch EC: Nuclear translocation of NFkappaB is increased in dopaminergic neurons of patients with
parkinson disease. Proc Natl Acad Sci U S A 1997, 94(14):75317536.
Iwata A, Maruyama M, Kanazawa I, Nukina N: alpha-Synuclein
affects the MAPK pathway and accelerates cell death. J Biol Chem
2001, 276(48):45320-45329.
Wang X, Yan MH, Fujioka H, Liu J, Wilson-Delfosse A, Chen SG,
Perry G, Casadesus G, Zhu X: LRRK2 regulates mitochondrial
dynamics and function through direct interaction with DLP1. Hum
Mol Genet 2012, 21(9):1931-1944.
Niu J, Yu M, Wang C, Xu Z: Leucine-rich repeat kinase 2 disturbs
mitochondrial dynamics via Dynamin-like protein. J Neurochem
2012, 122(3):650-658.
Lee SB, Kim W, Lee S, Chung J: Loss of LRRK2/PARK8 induces
degeneration of dopaminergic neurons in Drosophila. Biochem
Biophys Res Commun 2007, 358(2):534-539.
Saha S, Guillily MD, Ferree A, Lanceta J, Chan D, Ghosh J, Hsu
CH, Segal L, Raghavan K, Matsumoto K et al: LRRK2 modulates
vulnerability to mitochondrial dysfunction in Caenorhabditis
elegans. J Neurosci 2009, 29(29):9210-9218.
Ng CH, Mok SZ, Koh C, Ouyang X, Fivaz ML, Tan EK, Dawson
VL, Dawson TM, Yu F, Lim KL: Parkin protects against LRRK2
G2019S mutant-induced dopaminergic neurodegeneration in
Drosophila. J Neurosci 2009, 29(36):11257-11262.
Mortiboys H, Johansen KK, Aasly JO, Bandmann O:
Mitochondrial impairment in patients with Parkinson disease with
the G2019S mutation in LRRK2. Neurology 2010, 75(22):20172020.
Cui J, Yu M, Niu J, Yue Z, Xu Z: Expression of Leucine-Rich
Repeat Kinase 2 (LRRK2) Inhibits the Processing of uMtCK to
Induce Cell Death in cell culture model system. Biosci Rep 2011.
Venderova K, Kabbach G, Abdel-Messih E, Zhang Y, Parks RJ,
Imai Y, Gehrke S, Ngsee J, Lavoie MJ, Slack RS et al: LeucineRich Repeat Kinase 2 interacts with Parkin, DJ-1 and PINK-1 in a
Drosophila melanogaster model of Parkinson's disease. Hum Mol
Genet 2009, 18(22):4390-4404.
Nishio K, Qiao S, Yamashita H: Characterization of the
differential expression of uncoupling protein 2 and ROS
289
289.
290.
291.
292.
293.
294.
295.
296.
297.
298.
299.
300.
301.
302.
production in differentiated mouse macrophage-cells (Mm1) and
the progenitor cells (M1). J Mol Histol 2005, 36(1-2):35-44.
Andres-Mateos E, Mejias R, Sasaki M, Li X, Lin BM, Biskup S,
Zhang L, Banerjee R, Thomas B, Yang L et al: Unexpected lack of
hypersensitivity in LRRK2 knock-out mice to MPTP (1-methyl-4phenyl-1,2,3,6-tetrahydropyridine). J Neurosci 2009, 29(50):1584615850.
Bradley JL, Homayoun S, Hart PE, Schapira AH, Cooper JM:
Role of oxidative damage in Friedreich's ataxia. Neurochem Res
2004, 29(3):561-567.
Tabrizi SJ, Cleeter MW, Xuereb J, Taanman JW, Cooper JM,
Schapira AH: Biochemical abnormalities and excitotoxicity in
Huntington's disease brain. Ann Neurol 1999, 45(1):25-32.
Chau KY, Cooper JM, Schapira AH: Rasagiline protects against
alpha-synuclein induced sensitivity to oxidative stress in
dopaminergic cells. Neurochem Int 2010, 57(5):525-529.
Chau KY, Korlipara LV, Cooper JM, Schapira AH: Protection
against paraquat and A53T alpha-synuclein toxicity by
cabergoline is partially mediated by dopamine receptors. J Neurol
Sci 2009, 278(1-2):44-53.
Smith PK, Krohn RI, Hermanson GT, Mallia AK, Gartner FH,
Provenzano MD, Fujimoto EK, Goeke NM, Olson BJ, Klenk DC:
Measurement of protein using bicinchoninic acid. Anal Biochem
1985, 150(1):76-85.
Schagger H: Blue-native gels to isolate protein complexes from
mitochondria. Methods Cell Biol 2001, 65:231-244.
Rahman S, Lake BD, Taanman JW, Hanna MG, Cooper JM,
Schapira
AH,
Leonard
JV:
Cytochrome
oxidase
immunohistochemistry: clues for genetic mechanisms. Brain 2000,
123 Pt 3:591-600.
Waltregny D, North B, Van Mellaert F, de Leval J, Verdin E,
Castronovo V: Screening of histone deacetylases (HDAC)
expression in human prostate cancer reveals distinct class I HDAC
profiles between epithelial and stromal cells. Eur J Histochem
2004, 48(3):273-290.
Waxman EA, Covy JP, Bukh I, Li X, Dawson TM, Giasson BI:
Leucine-rich repeat kinase 2 expression leads to aggresome
formation that is not associated with alpha-synuclein inclusions. J
Neuropathol Exp Neurol 2009, 68(7):785-796.
Lillie RD: Histopathologic technic and practical histochemistry, 3d
edn. New York: McGraw-Hill; 1965.
Graham JM: Isolation of mitochondria from tissues and cells by
differential centrifugation. Curr Protoc Cell Biol 2001, Chapter
3:Unit 3 3.
Hornig-Do HT, Gunther G, Bust M, Lehnartz P, Bosio A, Wiesner
RJ:
Isolation
of
functional
pure
mitochondria
by
superparamagnetic microbeads. Anal Biochem 2009, 389(1):1-5.
Pon LA, Schon EA, American Society for Cell Biology.:
Mitochondria, 2nd edn. San Diego, Calif.: Academic Press; 2007.
290
303.
304.
305.
306.
307.
308.
309.
310.
311.
312.
313.
314.
315.
316.
Kouri K, Duchen MR, Lemmens-Gruber R: Effects of beauvericin
on the metabolic state and ionic homeostasis of ventricular
myocytes of the guinea pig. Chem Res Toxicol 2005, 18(11):16611668.
Wilson-Fritch L, Burkart A, Bell G, Mendelson K, Leszyk J,
Nicoloro S, Czech M, Corvera S: Mitochondrial biogenesis and
remodeling during adipogenesis and in response to the insulin
sensitizer rosiglitazone. Mol Cell Biol 2003, 23(3):1085-1094.
Hynes J, Natoli E, Jr., Will Y: Fluorescent pH and oxygen probes
of the assessment of mitochondrial toxicity in isolated
mitochondria and whole cells. Curr Protoc Toxicol 2009, Chapter
2:Unit 2 16.
Zhdanov AV, Favre C, O'Flaherty L, Adam J, O'Connor R,
Pollard PJ, Papkovsky DB: Comparative bioenergetic assessment
of transformed cells using a cell energy budget platform. Integr
Biol (Camb) 2011, 3(11):1135-1142.
Gullbo J, Fryknas M, Rickardson L, Darcy P, Hagg M, Wickstrom
M, Hassan S, Westman G, Brnjic S, Nygren P et al: Phenotypebased drug screening in primary ovarian carcinoma cultures
identifies intracellular iron depletion as a promising strategy for
cancer treatment. Biochem Pharmacol 2011, 82(2):139-147.
Delves PJ: Roitt's essential immunology, 11th edn. Malden, Mass. ;
Oxford: Blackwell; 2006.
Michalak M, Corbett EF, Mesaeli N, Nakamura K, Opas M:
Calreticulin: one protein, one gene, many functions. Biochem J
1999, 344 Pt 2:281-292.
Chinopoulos C, Gerencser AA, Mandi M, Mathe K, Torocsik B,
Doczi J, Turiak L, Kiss G, Konrad C, Vajda S et al: Forward
operation of adenine nucleotide translocase during F0F1-ATPase
reversal: critical role of matrix substrate-level phosphorylation.
FASEB J 2010, 24(7):2405-2416.
Vercesi AE, Kowaltowski AJ, Grijalba MT, Meinicke AR, Castilho
RF: The role of reactive oxygen species in mitochondrial
permeability transition. Biosci Rep 1997, 17(1):43-52.
Rasmussen HN, Andersen AJ, Rasmussen UF: Optimization of
preparation of mitochondria from 25-100 mg skeletal muscle. Anal
Biochem 1997, 252(1):153-159.
Krauss S, Zhang CY, Lowell BB: The mitochondrial uncouplingprotein homologues. Nat Rev Mol Cell Biol 2005, 6(3):248-261.
Kajimoto K, Yamazaki N, Kataoka M, Terada H, Shinohara Y:
Identification of possible protein machinery involved in the
thermogenic function of brown adipose tissue. J Med Invest 2004,
51(1-2):20-28.
Zhang CY, Parton LE, Ye CP, Krauss S, Shen R, Lin CT, Porco
JA, Jr., Lowell BB: Genipin inhibits UCP2-mediated proton leak
and acutely reverses obesity- and high glucose-induced beta cell
dysfunction in isolated pancreatic islets. Cell Metab 2006, 3(6):417427.
Puigserver P, Spiegelman BM: Peroxisome proliferator-activated
receptor-gamma coactivator 1 alpha (PGC-1 alpha):
291
317.
318.
319.
320.
321.
322.
323.
324.
325.
326.
327.
328.
329.
transcriptional coactivator and metabolic regulator. Endocr Rev
2003, 24(1):78-90.
Vega RB, Harrison BC, Meadows E, Roberts CR, Papst PJ, Olson
EN, McKinsey TA: Protein kinases C and D mediate agonistdependent cardiac hypertrophy through nuclear export of histone
deacetylase 5. Mol Cell Biol 2004, 24(19):8374-8385.
Galter D, Westerlund M, Carmine A, Lindqvist E, Sydow O, Olson
L: LRRK2 expression linked to dopamine-innervated areas. Ann
Neurol 2006, 59(4):714-719.
Taymans JM, Van den Haute C, Baekelandt V: Distribution of
PINK1 and LRRK2 in rat and mouse brain. J Neurochem 2006,
98(3):951-961.
Giasson BI, Covy JP, Bonini NM, Hurtig HI, Farrer MJ,
Trojanowski JQ, Van Deerlin VM: Biochemical and pathological
characterization of Lrrk2. Ann Neurol 2006, 59(2):315-322.
Moehle MS, Webber PJ, Tse T, Sukar N, Standaert DG, DeSilva
TM, Cowell RM, West AB: LRRK2 inhibition attenuates
microglial inflammatory responses. J Neurosci 2012, 32(5):16021611.
Thevenet J, Pescini Gobert R, Hooft van Huijsduijnen R, Wiessner
C, Sagot YJ: Regulation of LRRK2 expression points to a
functional role in human monocyte maturation. PLoS One 2011,
6(6):e21519.
Dusonchet J, Kochubey O, Stafa K, Young SM, Jr., Zufferey R,
Moore DJ, Schneider BL, Aebischer P: A rat model of progressive
nigral neurodegeneration induced by the Parkinson's diseaseassociated G2019S mutation in LRRK2. J Neurosci 2011,
31(3):907-912.
Devine MJ, Kaganovich A, Ryten M, Mamais A, Trabzuni D,
Manzoni C, McGoldrick P, Chan D, Dillman A, Zerle J et al:
Pathogenic LRRK2 mutations do not alter gene expression in cell
model systems or human brain tissue. PLoS One 2011, 6(7):e22489.
Maekawa T, Kubo M, Yokoyama I, Ohta E, Obata F: Agedependent and cell-population-restricted LRRK2 expression in
normal mouse spleen. Biochem Biophys Res Commun 2010,
392(3):431-435.
Milosevic J, Schwarz SC, Ogunlade V, Meyer AK, Storch A,
Schwarz J: Emerging role of LRRK2 in human neural progenitor
cell cycle progression, survival and differentiation. Mol
Neurodegener 2009, 4:25.
Chan D, Citro A, Cordy JM, Shen GC, Wolozin B: Rac1 protein
rescues neurite retraction caused by G2019S leucine-rich repeat
kinase 2 (LRRK2). J Biol Chem 2011, 286(18):16140-16149.
Cho HJ, Liu G, Jin SM, Parisiadou L, Xie C, Yu J, Sun L, Ma B,
Ding J, Vancraenenbroeck R et al: MicroRNA-205 regulates the
expression of Parkinson's disease-related leucine-rich repeat
kinase 2 protein. Hum Mol Genet 2012.
Melrose HL, Dachsel JC, Behrouz B, Lincoln SJ, Yue M, Hinkle
KM, Kent CB, Korvatska E, Taylor JP, Witten L et al: Impaired
dopaminergic neurotransmission and microtubule-associated
292
330.
331.
332.
333.
334.
335.
336.
337.
338.
339.
340.
341.
342.
343.
protein tau alterations in human LRRK2 transgenic mice.
Neurobiol Dis 2010, 40(3):503-517.
Maekawa T, Mori S, Sasaki Y, Miyajima T, Azuma S, Ohta E,
Obata F: The I2020T Leucine-rich repeat kinase 2 transgenic
mouse exhibits impaired locomotive ability accompanied by
dopaminergic neuron abnormalities. Mol Neurodegener 2012, 7:15.
Sharma S, Bandopadhyay R, Lashley T, Renton AE, Kingsbury
AE, Kumaran R, Kallis C, Vilarino-Guell C, O'Sullivan SS, Lees
AJ et al: LRRK2 expression in idiopathic and G2019S positive
Parkinson's disease subjects: a morphological and quantitative
study. Neuropathol Appl Neurobiol 2011, 37(7):777-790.
Alegre-Abarrategui J, Wade-Martins R: Parkinson disease,
LRRK2 and the endocytic-autophagic pathway. Autophagy 2009,
5(8):1208-1210.
Kett LR, Boassa D, Ho CC, Rideout HJ, Hu J, Terada M, Ellisman
M, Dauer WT: LRRK2 Parkinson disease mutations enhance its
microtubule association. Hum Mol Genet 2012, 21(4):890-899.
Tsuruta F, Sunayama J, Mori Y, Hattori S, Shimizu S, Tsujimoto
Y, Yoshioka K, Masuyama N, Gotoh Y: JNK promotes Bax
translocation to mitochondria through phosphorylation of 14-3-3
proteins. EMBO J 2004, 23(8):1889-1899.
Schwartz MP, Matouschek A: The dimensions of the protein
import channels in the outer and inner mitochondrial membranes.
Proc Natl Acad Sci U S A 1999, 96(23):13086-13090.
Lin W, Kang UJ: Structural determinants of PINK1 topology and
dual subcellular distribution. BMC Cell Biol 2010, 11:90.
Badugu R, Garcia M, Bondada V, Joshi A, Geddes JW: N
terminus of calpain 1 is a mitochondrial targeting sequence. J Biol
Chem 2008, 283(6):3409-3417.
Melrose H: Update on the functional biology of Lrrk2. Future
Neurol 2008, 3(6):669-681.
Hermanson SB, Carlson CB, Riddle SM, Zhao J, Vogel KW,
Nichols RJ, Bi K: Screening for novel LRRK2 inhibitors using a
high-throughput TR-FRET cellular assay for LRRK2 Ser935
phosphorylation. PLoS One 2012, 7(8):e43580.
Parker WD, Jr., Boyson SJ, Parks JK: Abnormalities of the
electron transport chain in idiopathic Parkinson's disease. Ann
Neurol 1989, 26(6):719-723.
Tieu K, Ischiropoulos H, Przedborski S: Nitric oxide and reactive
oxygen species in Parkinson's disease. IUBMB Life 2003,
55(6):329-335.
Tieu K, Perier C, Caspersen C, Teismann P, Wu DC, Yan SD,
Naini A, Vila M, Jackson-Lewis V, Ramasamy R et al: D-betahydroxybutyrate rescues mitochondrial respiration and mitigates
features of Parkinson disease. J Clin Invest 2003, 112(6):892-901.
Grunewald A, Gegg ME, Taanman JW, King RH, Kock N, Klein
C, Schapira AH: Differential effects of PINK1 nonsense and
missense mutations on mitochondrial function and morphology.
Exp Neurol 2009, 219(1):266-273.
293
344.
345.
346.
347.
348.
349.
350.
351.
352.
353.
354.
355.
356.
357.
Gautier CA, Kitada T, Shen J: Loss of PINK1 causes
mitochondrial functional defects and increased sensitivity to
oxidative stress. Proc Natl Acad Sci U S A 2008, 105(32):1136411369.
Vincent A, Briggs L, Chatwin GF, Emery E, Tomlins R, Oswald
M, Middleton CA, Evans GJ, Sweeney ST, Elliott CJ: parkininduced defects in neurophysiology and locomotion are generated
by metabolic dysfunction and not oxidative stress. Hum Mol Genet
2012, 21(8):1760-1769.
Grunewald A, Voges L, Rakovic A, Kasten M, Vandebona H,
Hemmelmann C, Lohmann K, Orolicki S, Ramirez A, Schapira
AH et al: Mutant Parkin impairs mitochondrial function and
morphology in human fibroblasts. PLoS One 2010, 5(9):e12962.
Mortiboys H, Thomas KJ, Koopman WJ, Klaffke S, Abou-Sleiman
P, Olpin S, Wood NW, Willems PH, Smeitink JA, Cookson MR et
al: Mitochondrial function and morphology are impaired in
parkin-mutant fibroblasts. Ann Neurol 2008, 64(5):555-565.
Gautier CA, Giaime E, Caballero E, Nunez L, Song Z, Chan D,
Villalobos C, Shen J: Regulation of mitochondrial permeability
transition pore by PINK1. Mol Neurodegener 2012, 7:22.
Jastroch M, Divakaruni AS, Mookerjee S, Treberg JR, Brand
MD: Mitochondrial proton and electron leaks. Essays Biochem
2010, 47:53-67.
Krauss S, Zhang CY, Lowell BB: A significant portion of
mitochondrial proton leak in intact thymocytes depends on
expression of UCP2. Proc Natl Acad Sci U S A 2002, 99(1):118-122.
Rolfe DF, Brand MD: Contribution of mitochondrial proton leak
to skeletal muscle respiration and to standard metabolic rate. Am J
Physiol 1996, 271(4 Pt 1):C1380-1389.
Rasola A, Bernardi P: The mitochondrial permeability transition
pore and its involvement in cell death and in disease pathogenesis.
Apoptosis 2007, 12(5):815-833.
Vaseva AV, Marchenko ND, Ji K, Tsirka SE, Holzmann S, Moll
UM: p53 opens the mitochondrial permeability transition pore to
trigger necrosis. Cell 2012, 149(7):1536-1548.
Rousset S, Alves-Guerra MC, Mozo J, Miroux B, CassardDoulcier AM, Bouillaud F, Ricquier D: The biology of
mitochondrial uncoupling proteins. Diabetes 2004, 53 Suppl
1:S130-135.
Andrews ZB, Horvath TL: Uncoupling protein-2 regulates lifespan
in mice. Am J Physiol Endocrinol Metab 2009, 296(4):E621-627.
Casteilla L, Blondel O, Klaus S, Raimbault S, Diolez P, Moreau F,
Bouillaud F, Ricquier D: Stable expression of functional
mitochondrial uncoupling protein in Chinese hamster ovary cells.
Proc Natl Acad Sci U S A 1990, 87(13):5124-5128.
Ho PW, Ho JW, Tse HM, So DH, Yiu DC, Liu HF, Chan KH,
Kung MH, Ramsden DB, Ho SL: Uncoupling protein-4 (UCP4)
increases ATP supply by interacting with mitochondrial Complex
II in neuroblastoma cells. PLoS One 2012, 7(2):e32810.
294
358.
359.
360.
361.
362.
363.
364.
365.
366.
367.
368.
369.
370.
Hoang T, Smith MD, Jelokhani-Niaraki M: Toward
understanding the mechanism of ion transport activity of neuronal
uncoupling proteins UCP2, UCP4, and UCP5. Biochemistry 2012,
51(19):4004-4014.
Cannon B, Nedergaard J: Brown adipose tissue: function and
physiological significance. Physiol Rev 2004, 84(1):277-359.
Chan CB, De Leo D, Joseph JW, McQuaid TS, Ha XF, Xu F,
Tsushima RG, Pennefather PS, Salapatek AM, Wheeler MB:
Increased uncoupling protein-2 levels in beta-cells are associated
with impaired glucose-stimulated insulin secretion: mechanism of
action. Diabetes 2001, 50(6):1302-1310.
Mori S, Yoshizuka N, Takizawa M, Takema Y, Murase T,
Tokimitsu I, Saito M: Expression of uncoupling proteins in human
skin and skin-derived cells. J Invest Dermatol 2008, 128(8):18941900.
Lengacher S, Magistretti PJ, Pellerin L: Quantitative rt-PCR
analysis of uncoupling protein isoforms in mouse brain cortex:
methodological optimization and comparison of expression with
brown adipose tissue and skeletal muscle. J Cereb Blood Flow
Metab 2004, 24(7):780-788.
Wu Z, Zhang J, Zhao B: Superoxide anion regulates the
mitochondrial free Ca2+ through uncoupling proteins. Antioxid
Redox Signal 2009, 11(8):1805-1818.
Liu Y, Jiang H, Xing FQ, Huang WJ, Mao LH, He LY:
Uncoupling protein 2 expression affects androgen synthesis in
polycystic ovary syndrome. Endocrine 2012.
Madsen L, Pedersen LM, Lillefosse HH, Fjaere E, Bronstad I, Hao
Q, Petersen RK, Hallenborg P, Ma T, De Matteis R et al: UCP1
induction during recruitment of brown adipocytes in white adipose
tissue is dependent on cyclooxygenase activity. PLoS One 2010,
5(6):e11391.
Emilsson V, O'Dowd J, Wang S, Liu YL, Sennitt M, Heyman R,
Cawthorne MA: The effects of rexinoids and rosiglitazone on body
weight and uncoupling protein isoform expression in the Zucker
fa/fa rat. Metabolism 2000, 49(12):1610-1615.
Li Y, Maedler K, Shu L, Haataja L: UCP-2 and UCP-3 proteins
are differentially regulated in pancreatic beta-cells. PLoS One
2008, 3(1):e1397.
Rupprecht A, Brauer AU, Smorodchenko A, Goyn J, Hilse KE,
Shabalina IG, Infante-Duarte C, Pohl EE: Quantification of
uncoupling protein 2 reveals its main expression in immune cells
and selective up-regulation during T-cell proliferation. PLoS One
2012, 7(8):e41406.
Mailloux RJ, Adjeitey CN, Harper ME: Genipin-induced
inhibition of uncoupling protein-2 sensitizes drug-resistant cancer
cells to cytotoxic agents. PLoS One 2010, 5(10):e13289.
Huntgeburth M, Tiemann K, Shahverdyan R, Schluter KD,
Schreckenberg R, Gross ML, Modersheim S, Caglayan E, MullerEhmsen J, Ghanem A et al: Transforming growth factor beta(1)
oppositely regulates the hypertrophic and contractile response to
295
371.
372.
373.
374.
375.
376.
377.
378.
379.
380.
381.
beta-adrenergic stimulation in the heart. PLoS One 2011,
6(11):e26628.
Kim BC, Kim HG, Lee SA, Lim S, Park EH, Kim SJ, Lim CJ:
Genipin-induced apoptosis in hepatoma cells is mediated by
reactive oxygen species/c-Jun NH2-terminal kinase-dependent
activation of mitochondrial pathway. Biochem Pharmacol 2005,
70(9):1398-1407.
Diao J, Allister EM, Koshkin V, Lee SC, Bhattacharjee A, Tang C,
Giacca A, Chan CB, Wheeler MB: UCP2 is highly expressed in
pancreatic alpha-cells and influences secretion and survival. Proc
Natl Acad Sci U S A 2008, 105(33):12057-12062.
Wu Z, Puigserver P, Andersson U, Zhang C, Adelmant G, Mootha
V, Troy A, Cinti S, Lowell B, Scarpulla RC et al: Mechanisms
controlling mitochondrial biogenesis and respiration through the
thermogenic coactivator PGC-1. Cell 1999, 98(1):115-124.
Lehman JJ, Barger PM, Kovacs A, Saffitz JE, Medeiros DM, Kelly
DP: Peroxisome proliferator-activated
receptor gamma
coactivator-1 promotes cardiac mitochondrial biogenesis. J Clin
Invest 2000, 106(7):847-856.
Puigserver P, Wu Z, Park CW, Graves R, Wright M, Spiegelman
BM: A cold-inducible coactivator of nuclear receptors linked to
adaptive thermogenesis. Cell 1998, 92(6):829-839.
Evans MJ, Scarpulla RC: NRF-1: a trans-activator of nuclearencoded respiratory genes in animal cells. Genes Dev 1990,
4(6):1023-1034.
Virbasius CA, Virbasius JV, Scarpulla RC: NRF-1, an activator
involved in nuclear-mitochondrial interactions, utilizes a new
DNA-binding domain conserved in a family of developmental
regulators. Genes Dev 1993, 7(12A):2431-2445.
Nagase I, Yoshida S, Canas X, Irie Y, Kimura K, Yoshida T, Saito
M: Up-regulation of uncoupling protein 3 by thyroid hormone,
peroxisome proliferator-activated receptor ligands and 9-cis
retinoic acid in L6 myotubes. FEBS Lett 1999, 461(3):319-322.
Son C, Hosoda K, Matsuda J, Fujikura J, Yonemitsu S, Iwakura
H, Masuzaki H, Ogawa Y, Hayashi T, Itoh H et al: Up-regulation
of uncoupling protein 3 gene expression by fatty acids and agonists
for PPARs in L6 myotubes. Endocrinology 2001, 142(10):41894194.
Brun S, Carmona MC, Mampel T, Vinas O, Giralt M, Iglesias R,
Villarroya F: Activators of peroxisome proliferator-activated
receptor-alpha induce the expression of the uncoupling protein-3
gene in skeletal muscle: a potential mechanism for the lipid intakedependent activation of uncoupling protein-3 gene expression at
birth. Diabetes 1999, 48(6):1217-1222.
Chevillotte E, Rieusset J, Roques M, Desage M, Vidal H: The
regulation of uncoupling protein-2 gene expression by omega-6
polyunsaturated fatty acids in human skeletal muscle cells involves
multiple pathways, including the nuclear receptor peroxisome
proliferator-activated receptor beta. J Biol Chem 2001,
276(14):10853-10860.
296
382.
383.
384.
385.
386.
387.
388.
389.
390.
391.
392.
393.
Solanes G, Pedraza N, Iglesias R, Giralt M, Villarroya F:
Functional relationship between MyoD and peroxisome
proliferator-activated receptor-dependent regulatory pathways in
the control of the human uncoupling protein-3 gene transcription.
Mol Endocrinol 2003, 17(10):1944-1958.
Li X, Wang QJ, Pan N, Lee S, Zhao Y, Chait BT, Yue Z:
Phosphorylation-dependent 14-3-3 binding to LRRK2 is impaired
by common mutations of familial Parkinson's disease. PLoS One
2011, 6(3):e17153.
Ha CH, Kim JY, Zhao J, Wang W, Jhun BS, Wong C, Jin ZG:
PKA phosphorylates histone deacetylase 5 and prevents its nuclear
export, leading to the inhibition of gene transcription and
cardiomyocyte hypertrophy. Proc Natl Acad Sci U S A 2010,
107(35):15467-15472.
Sucharov CC, Langer S, Bristow M, Leinwand L: Shuttling of
HDAC5 in H9C2 cells regulates YY1 function through
CaMKIV/PKD and PP2A. Am J Physiol Cell Physiol 2006,
291(5):C1029-1037.
McKinsey TA, Zhang CL, Olson EN: Identification of a signalresponsive nuclear export sequence in class II histone deacetylases.
Mol Cell Biol 2001, 21(18):6312-6321.
Gurvich N, Tsygankova OM, Meinkoth JL, Klein PS: Histone
deacetylase is a target of valproic acid-mediated cellular
differentiation. Cancer Res 2004, 64(3):1079-1086.
Baillie G, MacKenzie SJ, Houslay MD: Phorbol 12-myristate 13acetate triggers the protein kinase A-mediated phosphorylation
and activation of the PDE4D5 cAMP phosphodiesterase in human
aortic smooth muscle cells through a route involving extracellular
signal regulated kinase (ERK). Mol Pharmacol 2001, 60(5):11001111.
Yang TT, Xiong Q, Enslen H, Davis RJ, Chow CW:
Phosphorylation of NFATc4 by p38 mitogen-activated protein
kinases. Mol Cell Biol 2002, 22(11):3892-3904.
Chang CR, Blackstone C: Dynamic regulation of mitochondrial
fission through modification of the dynamin-related protein Drp1.
Ann N Y Acad Sci 2010, 1201:34-39.
Criscuolo C, De Rosa A, Guacci A, Simons EJ, Breedveld GJ,
Peluso S, Volpe G, Filla A, Oostra BA, Bonifati V et al: The
LRRK2 R1441C mutation is more frequent than G2019S in
Parkinson's disease patients from southern Italy. Mov Disord 2011,
26(9):1733-1736.
Nabben M, Hoeks J: Mitochondrial uncoupling protein 3 and its
role in cardiac- and skeletal muscle metabolism. Physiol Behav
2008, 94(2):259-269.
Mattson MP, Goodman Y, Luo H, Fu W, Furukawa K: Activation
of NF-kappaB protects hippocampal neurons against oxidative
stress-induced apoptosis: evidence for induction of manganese
superoxide dismutase and suppression of peroxynitrite production
and protein tyrosine nitration. J Neurosci Res 1997, 49(6):681-697.
297
394.
395.
396.
397.
398.
399.
400.
401.
402.
403.
404.
405.
406.
407.
Lee FY, Li Y, Zhu H, Yang S, Lin HZ, Trush M, Diehl AM:
Tumor necrosis factor increases mitochondrial oxidant production
and induces expression of uncoupling protein-2 in the regenerating
mice [correction of rat] liver. Hepatology 1999, 29(3):677-687.
Kim B, Yang MS, Choi D, Kim JH, Kim HS, Seol W, Choi S, Jou I,
Kim EY, Joe EH: Impaired inflammatory responses in murine
Lrrk2-knockdown brain microglia. PLoS One 2012, 7(4):e34693.
Azzu V, Mookerjee SA, Brand MD: Rapid turnover of
mitochondrial uncoupling protein 3. Biochem J 2010, 426(1):13-17.
Azzu V, Jastroch M, Divakaruni AS, Brand MD: The regulation
and turnover of mitochondrial uncoupling proteins. Biochim
Biophys Acta 2010, 1797(6-7):785-791.
Chen X, Wang K, Chen J, Guo J, Yin Y, Cai X, Guo X, Wang G,
Yang R, Zhu L et al: In vitro evidence suggests that miR-133amediated regulation of uncoupling protein 2 (UCP2) is an
indispensable step in myogenic differentiation. J Biol Chem 2009,
284(8):5362-5369.
Mailloux RJ, Seifert EL, Bouillaud F, Aguer C, Collins S, Harper
ME: Glutathionylation acts as a control switch for uncoupling
proteins UCP2 and UCP3. J Biol Chem 2011, 286(24):21865-21875.
Breen EP, Gouin SG, Murphy AF, Haines LR, Jackson AM,
Pearson TW, Murphy PV, Porter RK: On the mechanism of
mitochondrial uncoupling protein 1 function. J Biol Chem 2006,
281(4):2114-2119.
Twig G, Elorza A, Molina AJ, Mohamed H, Wikstrom JD, Walzer
G, Stiles L, Haigh SE, Katz S, Las G et al: Fission and selective
fusion govern mitochondrial segregation and elimination by
autophagy. EMBO J 2008, 27(2):433-446.
Yang Y, Ouyang Y, Yang L, Beal MF, McQuibban A, Vogel H, Lu
B: Pink1 regulates mitochondrial dynamics through interaction
with the fission/fusion machinery. Proc Natl Acad Sci U S A 2008,
105(19):7070-7075.
Wood-Kaczmar A, Gandhi S, Yao Z, Abramov AY, Miljan EA,
Keen G, Stanyer L, Hargreaves I, Klupsch K, Deas E et al: PINK1
is necessary for long term survival and mitochondrial function in
human dopaminergic neurons. PLoS One 2008, 3(6):e2455.
Clark IE, Dodson MW, Jiang C, Cao JH, Huh JR, Seol JH, Yoo
SJ, Hay BA, Guo M: Drosophila pink1 is required for
mitochondrial function and interacts genetically with parkin.
Nature 2006, 441(7097):1162-1166.
Park J, Lee SB, Lee S, Kim Y, Song S, Kim S, Bae E, Kim J, Shong
M, Kim JM et al: Mitochondrial dysfunction in Drosophila PINK1
mutants is complemented by parkin. Nature 2006, 441(7097):11571161.
Lyamzaev KG, Izyumov DS, Avetisyan AV, Yang F, Pletjushkina
OY, Chernyak BV: Inhibition of mitochondrial bioenergetics: the
effects on structure of mitochondria in the cell and on apoptosis.
Acta Biochim Pol 2004, 51(2):553-562.
Kieper N, Holmstrom KM, Ciceri D, Fiesel FC, Wolburg H,
Ziviani E, Whitworth AJ, Martins LM, Kahle PJ, Kruger R:
298
408.
409.
410.
411.
412.
413.
414.
415.
416.
417.
418.
Modulation of mitochondrial function and morphology by
interaction of Omi/HtrA2 with the mitochondrial fusion factor
OPA1. Exp Cell Res 2010, 316(7):1213-1224.
Nakamura T, Lipton SA: Redox regulation of mitochondrial
fission, protein misfolding, synaptic damage, and neuronal cell
death: potential implications for Alzheimer's and Parkinson's
diseases. Apoptosis 2010, 15(11):1354-1363.
Ichishita R, Tanaka K, Sugiura Y, Sayano T, Mihara K, Oka T:
An RNAi screen for mitochondrial proteins required to maintain
the morphology of the organelle in Caenorhabditis elegans. J
Biochem 2008, 143(4):449-454.
Barsoum MJ, Yuan H, Gerencser AA, Liot G, Kushnareva Y,
Graber S, Kovacs I, Lee WD, Waggoner J, Cui J et al: Nitric
oxide-induced mitochondrial fission is regulated by dynaminrelated GTPases in neurons. EMBO J 2006, 25(16):3900-3911.
Heeman B, Van den Haute C, Aelvoet SA, Valsecchi F, Rodenburg
RJ, Reumers V, Debyser Z, Callewaert G, Koopman WJ, Willems
PH et al: Depletion of PINK1 affects mitochondrial metabolism,
calcium homeostasis and energy maintenance. J Cell Sci 2011,
124(Pt 7):1115-1125.
Shavali S, Brown-Borg HM, Ebadi M, Porter J: Mitochondrial
localization of alpha-synuclein protein in alpha-synuclein
overexpressing cells. Neurosci Lett 2008, 439(2):125-128.
Plun-Favreau H, Burchell VS, Holmstrom KM, Yao Z, Deas E,
Cain K, Fedele V, Moisoi N, Campanella M, Miguel Martins L et
al: HtrA2 deficiency causes mitochondrial uncoupling through the
F(1)F(0)-ATP synthase and consequent ATP depletion. Cell Death
Dis 2012, 3:e335.
Srivastava RA, Pinkosky SL, Filippov S, Hanselman JC, Cramer
CT, Newton RS: AMP-activated protein kinase: an emerging drug
target to regulate imbalances in lipid and carbohydrate
metabolism to treat cardio-metabolic diseases: Thematic Review
Series: New Lipid and Lipoprotein Targets for the Treatment of
Cardiometabolic Diseases. J Lipid Res 2012, 53(12):2490-2514.
Lee JY, Takahashi N, Yasubuchi M, Kim YI, Hashizaki H, Kim
MJ, Sakamoto T, Goto T, Kawada T: Triiodothyronine induces
UCP-1 expression and mitochondrial biogenesis in human
adipocytes. Am J Physiol Cell Physiol 2012, 302(2):C463-472.
Rodriguez de la Concepcion ML, Yubero P, Domingo JC, Iglesias
R, Domingo P, Villarroya F, Giralt M: Reverse transcriptase
inhibitors alter uncoupling protein-1 and mitochondrial biogenesis
in brown adipocytes. Antivir Ther 2005, 10(4):515-526.
Patane G, Anello M, Piro S, Vigneri R, Purrello F, Rabuazzo AM:
Role of ATP production and uncoupling protein-2 in the insulin
secretory defect induced by chronic exposure to high glucose or
free fatty acids and effects of peroxisome proliferator-activated
receptor-gamma inhibition. Diabetes 2002, 51(9):2749-2756.
Han DH, Nolte LA, Ju JS, Coleman T, Holloszy JO, Semenkovich
CF: UCP-mediated energy depletion in skeletal muscle increases
glucose transport despite lipid accumulation and mitochondrial
299
419.
420.
421.
422.
423.
424.
425.
426.
427.
428.
429.
430.
431.
dysfunction. Am J Physiol Endocrinol Metab 2004, 286(3):E347353.
Trenker M, Malli R, Fertschai I, Levak-Frank S, Graier WF:
Uncoupling proteins 2 and 3 are fundamental for mitochondrial
Ca2+ uniport. Nat Cell Biol 2007, 9(4):445-452.
Yagi T, Kosakai A, Ito D, Okada Y, Akamatsu W, Nihei Y,
Nabetani A, Ishikawa F, Arai Y, Hirose N et al: Establishment of
induced pluripotent stem cells from centenarians for
neurodegenerative disease research. PLoS One 2012, 7(7):e41572.
Sanchez-Danes A, Richaud-Patin Y, Carballo-Carbajal I, JimenezDelgado S, Caig C, Mora S, Di Guglielmo C, Ezquerra M, Patel B,
Giralt A et al: Disease-specific phenotypes in dopamine neurons
from human iPS-based models of genetic and sporadic Parkinson's
disease. EMBO Mol Med 2012, 4(5):380-395.
Gomez-Suaga P, Luzon-Toro B, Churamani D, Zhang L, BloorYoung D, Patel S, Woodman PG, Churchill GC, Hilfiker S:
Leucine-rich repeat kinase 2 regulates autophagy through a
calcium-dependent pathway involving NAADP. Hum Mol Genet
2012, 21(3):511-525.
Kuballa P, Nolte WM, Castoreno AB, Xavier RJ: Autophagy and
the immune system. Annu Rev Immunol 2012, 30:611-646.
Lee JA: Neuronal autophagy: a housekeeper or a fighter in
neuronal cell survival? Exp Neurobiol 2012, 21(1):1-8.
Dando I, Fiorini C, Pozza ED, Padroni C, Costanzo C, Palmieri M,
Donadelli M: UCP2 inhibition triggers ROS-dependent nuclear
translocation of GAPDH and autophagic cell death in pancreatic
adenocarcinoma cells. Biochim Biophys Acta 2012.
Okatsu K, Oka T, Iguchi M, Imamura K, Kosako H, Tani N,
Kimura M, Go E, Koyano F, Funayama M et al: PINK1
autophosphorylation upon membrane potential dissipation is
essential for Parkin recruitment to damaged mitochondria. Nat
Commun 2012, 3:1016.
Gandhi PN, Wang X, Zhu X, Chen SG, Wilson-Delfosse AL: The
Roc domain of leucine-rich repeat kinase 2 is sufficient for
interaction with microtubules. J Neurosci Res 2008, 86(8):17111720.
Mattson MP, Liu D: Mitochondrial potassium channels and
uncoupling proteins in synaptic plasticity and neuronal cell death.
Biochem Biophys Res Commun 2003, 304(3):539-549.
Cousin MA, Robinson PJ: Ca(2+) influx inhibits dynamin and
arrests synaptic vesicle endocytosis at the active zone. J Neurosci
2000, 20(3):949-957.
Marks B, McMahon HT: Calcium triggers calcineurin-dependent
synaptic vesicle recycling in mammalian nerve terminals. Curr Biol
1998, 8(13):740-749.
Llorens-Martin M, Lopez-Domenech G, Soriano E, Avila J:
GSK3beta is involved in the relief of mitochondria pausing in a
Tau-dependent manner. PLoS One 2011, 6(11):e27686.
300
432.
433.
434.
435.
436.
437.
438.
439.
440.
441.
442.
Maldonado EN, Lemasters JJ: Warburg revisited: regulation of
mitochondrial metabolism by voltage-dependent anion channels in
cancer cells. J Pharmacol Exp Ther 2012, 342(3):637-641.
Sheng ZH, Cai Q: Mitochondrial transport in neurons: impact on
synaptic homeostasis and neurodegeneration. Nat Rev Neurosci
2012, 13(2):77-93.
Luft R, Ikkos D, Palmieri G, Ernster L, Afzelius B: A case of
severe hypermetabolism of nonthyroid origin with a defect in the
maintenance of mitochondrial respiratory control: a correlated
clinical, biochemical, and morphological study. J Clin Invest 1962,
41:1776-1804.
Sjostrand FS: Molecular pathology of Luft disease and structure
and function of mitochondria. J Submicrosc Cytol Pathol 1999,
31(1):41-50.
Ritz B, Rhodes SL, Qian L, Schernhammer E, Olsen JH, Friis S:
L-type calcium channel blockers and Parkinson disease in
Denmark. Ann Neurol 2010, 67(5):600-606.
Andrews ZB, Horvath B, Barnstable CJ, Elsworth J, Yang L, Beal
MF, Roth RH, Matthews RT, Horvath TL: Uncoupling protein-2
is critical for nigral dopamine cell survival in a mouse model of
Parkinson's disease. J Neurosci 2005, 25(1):184-191.
Conti B, Sugama S, Lucero J, Winsky-Sommerer R, Wirz SA,
Maher P, Andrews Z, Barr AM, Morale MC, Paneda C et al:
Uncoupling protein 2 protects dopaminergic neurons from acute
1,2,3,6-methyl-phenyl-tetrahydropyridine toxicity. J Neurochem
2005, 93(2):493-501.
Islam R, Yang L, Sah M, Kannan K, Anamani D, Vijayan C,
Kwok J, Cantino ME, Beal MF, Fridell YW: A neuroprotective
role of the human uncoupling protein 2 (hUCP2) in a Drosophila
Parkinson's disease model. Neurobiol Dis 2012, 46(1):137-146.
Sanchez-Blanco A, Fridell YW, Helfand SL: Involvement of
Drosophila uncoupling protein 5 in metabolism and aging.
Genetics 2006, 172(3):1699-1710.
Stockl P, Zankl C, Hutter E, Unterluggauer H, Laun P, Heeren G,
Bogengruber E, Herndler-Brandstetter D, Breitenbach M, JansenDurr P: Partial uncoupling of oxidative phosphorylation induces
premature senescence in human fibroblasts and yeast mother cells.
Free Radic Biol Med 2007, 43(6):947-958.
Desquiret V, Loiseau D, Jacques C, Douay O, Malthiery Y, Ritz P,
Roussel D: Dinitrophenol-induced mitochondrial uncoupling in
vivo triggers respiratory adaptation in HepG2 cells. Biochim
Biophys Acta 2006, 1757(1):21-30.
301
Appendix
302
303
304
305
Download